abordagem quali-quantitativa e funcional de vegetação campestre ...

156
2 UNIVERSIDADE FEDERAL DO RIO GRANDE DO SUL INSTITUTO DE BIOCIÊNCIAS PROGRAMA DE PÓS-GRADUAÇÃO EM BOTÂNICA ABORDAGEM QUALI-QUANTITATIVA E FUNCIONAL DE VEGETAÇÃO CAMPESTRE NOS BIOMAS PAMPA E MATA ATLÂNTICA Pedro Maria de Abreu Ferreira Orientação: Profa. Dra. Ilsi Iob Boldrini Co-Orientação: Prof. Dr. Gerhard Ernst Overbeck Tese apresentada ao Programa de Pós-Graduação em Botânica da Universidade Federal do Rio Grande do Sul como um dos pré-requisitos para a obtenção do título de doutor em Botânica Área de concentração em Ecologia Vegetal. Porto Alegre, 15 de Março de 2014

Transcript of abordagem quali-quantitativa e funcional de vegetação campestre ...

Page 1: abordagem quali-quantitativa e funcional de vegetação campestre ...

2

UNIVERSIDADE FEDERAL DO RIO GRANDE DO SUL

INSTITUTO DE BIOCIÊNCIAS

PROGRAMA DE PÓS-GRADUAÇÃO EM BOTÂNICA

ABORDAGEM QUALI-QUANTITATIVA E

FUNCIONAL DE VEGETAÇÃO CAMPESTRE NOS

BIOMAS PAMPA E MATA ATLÂNTICA

Pedro Maria de Abreu Ferreira

Orientação: Profa. Dra. Ilsi Iob Boldrini

Co-Orientação: Prof. Dr. Gerhard Ernst Overbeck

Tese apresentada ao Programa de Pós-Graduação em Botânica da

Universidade Federal do Rio Grande do Sul como um dos pré-requisitos para

a obtenção do título de doutor em Botânica – Área de concentração em

Ecologia Vegetal.

Porto Alegre, 15 de Março de 2014

Page 2: abordagem quali-quantitativa e funcional de vegetação campestre ...

3

Índice

1. Agradecimentos …........................................................................................................................ 4

2. Introduçao geral ............................................................................................................................ 7

3. Capítulo 1 – Differences in grassland vegetation from two South Brazilian biomes and

implications for conservation ................................................................................................ 23

4. Capítulo 2 – Plant life forms revisited: are classic systems really applicable in all ecosystems?

…………………………………………………………………………………………........ 54

5. Capítulo 3 – Exploring the relationships between habitat structure and arthropod diversity in

South Brazilian grasslands: a functional perspective …………………………………........ 94

6. Capítulo 4 – Long-term ecological research in subtropical grasslands: results from a four-year

monitoring of different management practices in Southern Brazil …………..………....... 122

7. Considerações finais ................................................................................................................. 156

Page 3: abordagem quali-quantitativa e funcional de vegetação campestre ...

4

“O começo de todas as ciências é o espanto de as coisas serem o que são”.

Aristóteles

“Eu acredito demais na sorte. E tenho constatado que, quanto mais duro eu

trabalho, mais sorte eu tenho”.

Thomas Jefferson

Page 4: abordagem quali-quantitativa e funcional de vegetação campestre ...

5

Agradecimentos

É injusto que uma tese de doutorado só tenha um nome na capa. Ela é o produto de

incontáveis parcerias, discussões, almoços no RU e mesas de bar. Dito isto, vamos aos

agradecimentos.

Antes de mais nada agradeço a todos em geral que me aguentaram durante estes quatro

anos (pra alguns bem mais do que isso). Como já me disseram, eu ‘ou falo demais ou de menos, e

sou muito teimoso pra tudo, até para negar esse fato’. Eu obviamente discordo, mas obrigado por

me aturarem de qualquer maneira.

Agradeço à minha família: Vera, Maria Eugênia, Jimi e Tibi, por tudo (e já aproveito para

me desculpar por qualquer coisa também). A base da formação se dá, sem dúvida, em casa. No meu

caso não sei se isso é bom, mas que é verdade, é. Agradeço também à Grasi, que já é da família

também, por ser exatamente como ela é.

Amigos são a família que nós escolhemos. Obrigado a todos os amigos, os antigos, os

novos, os de longe e os que moram na esquina e tem mais facilidade no processo de atrasar o meu

trabalho em geral. Já escrevi isso em outro lugar, mas escrevo de novo: sem amigos eu seria muito

mais produtivo mas muito menos feliz. Agradeço em especial à ‘gurizada’: Robson, Dêniel,

Marcelo e Alecrim, e às respectivas cônjuges Edi, Carol, Daia e Giovana. Sem vocês eu teria mais

fígado e menos barriga.

Também não posso deixar de agradecer aos amigos do nosso laboratório, o LEVCamp, e os

do ECOQUA, do qual me orgulho em ser membro-visitante-agregado. Não vou citar nomes aqui,

vocês sabem quem são, e provavelmente eu esqueceria de alguém porque estou escrevendo isso

algumas horas antes de entregar a tese.

Considerando o processo de coleta de dados do doutorado, os maiores agradecimentos vão

para os proprietários rurais que permitiram que o trabalho fosse realizado em suas propriedades.

Sem vocês simplesmente não haveria trabalho. Apesar de parecer simples, abrir a sua casa para

Page 5: abordagem quali-quantitativa e funcional de vegetação campestre ...

6

estranhos, ainda mais em um projeto que teoricamente não tem limite de tempo, não é algo trivial,

mas digno de agradecimento constante e irrestrito. A construção de pontes entre a sociedade e a

academia é fundamental. Desenvolver projetos de pesquisa em propriedades rurais é uma

oportunidade única para isso. Obrigado Cláudio, Júlio, Adauto e Paulo Rogério. Agradeço também

às equipes da SEMA de São Francisco de Paula (Daniel e todos os outros gestores e outros

functionários) e do Parque Nacional dos Aparados da Serra (Deonir e demais), que tornaram

possível a implantação dos três sítios nos Campos de Cima da Serra.

Finalmente, agadeço aos meus orientadores, Ilsi e Gerhard. Muito obrigado principalmente

por permitir (e até incentivar) minha participação em tantos outros projetos fora do âmbito restrito

da tese de doutorado. Obrigado também pelas discussões e pela orientação em geral. Deixo o

laboratório já com saudades antecipadas. Aliás, obrigado aos dois também pelo ambiente de

trabalho que temos no laboratório. Convivi por diferentes períodos de tempo em diversos outros

laboratórios da UFRGS e de outras instituições, e nenhum deles é tão divertido, etílico e

consequentemente produtivo como o nosso. Tive também vários orientadores extra-oficiais, tanto

professores quanto colegas. Agradeço em especial ao Valério, que possibilitou a minha participação

em diversos projetos e atividades junto ao ECOQUA.

Page 6: abordagem quali-quantitativa e funcional de vegetação campestre ...

7

Introdução Geral

As formações campestres em senso amplo (i.e., incluindo savanas, vegetação arbustiva e

tundra), cobrem 52,5 milhões de quilômetros quadrados, o que corresponde a 40,5% da superfície

da Terra, excluindo-se Groenlândia e Antártida (Suttie et al. 2005). Na América do Sul, as

formações campestres, em senso amplo, cobrem cerca de 3,5 milhões de quilômetros quadrados, ou

350 milhões de hectares (Burkart 1975). No Brasil, os campos naturais cobrem 13.656.000 ha

(IBGE 2006), se concentram na região sul do país e foram as formações dominantes na região

durante o Pleistoceno recente, sendo sua distribuição atual interpretada como um indício de um

clima anterior mais seco e frio (Behling 2002; Bredenkamp et al. 2002).

Os campos são as formações fitoecológicas predominantes no Rio Grande do Sul (RS),

cobrem 62,2% da superfície do Estado (Cordeiro & Hasenack 2009) e se distribuem em dois biomas

distintos: Pampa e Mata Atlântica (IBGE 2004; Boldrini 2009). No bioma Pampa são encontradas

as maiores extensões contínuas de campo natural no Estado, distribuídas predominantemente em

terras baixas de relevo suave-ondulado, enquanto que no bioma Mata Atlântica os campos

apresentam distribuição em mosaico com as Florestas Ombrófilas Densa e Mista,

predominantemente em regiões de altitudes e de relevo escarpado. Burkart (1975) classifica a

vegetação campestre sul-brasileira em dois tipos: os “Campos do Brasil Central”, que abrangem as

formações campestres do norte do RS, de Santa Catarina e do Paraná, e os “Campos do Uruguai e

sul do Brasil”, que incluem os campos da metade sul do Estado e sua continuidade florística no

Uruguai. A classificação proposta por Burkart (1975) coincide com a divisão atual dos biomas Mata

Atlântica e Pampa no RS.

Considerando as áreas utilizadas para atividade pecuária sobre campo natural como

fisionomias seminaturais, 68,62% da superfície do RS encontra-se convertida para uso humano, e

estima-se que a taxa de conversão de fisionomias naturais para antrópicas seja de ca. 1000 km2/ano

(Cordeiro & Hasenack 2009). Dos 31,38% de fisionomias naturais ou seminaturais que ainda

Page 7: abordagem quali-quantitativa e funcional de vegetação campestre ...

8

cobrem o Estado, 62,21% são compostas por formações campestres, fato que reflete o caráter

ambientalmente sustentável da histórica prática da pecuária extensiva sobre campo nativo (Cordeiro

& Hasenack 2009). Como apontado por Crawshaw et al. (2007), esta atividade, historicamente

conduzida nos campos do RS, é um dos poucos exemplos de viabilidade econômica e

sustentabilidade quando comparada à agricultura. Dentre os Estados que compõem a região Sul do

Brasil, a maior área de campo natural preservada encontra-se no RS, tendo em vista que nos Estados

do Paraná e Santa Catarina restam respectivamente 1.377 e 1.779 milhões de hectares de área

campestre natural (IBGE 2006).

Os campos no RS são formações que apresentam elevada riqueza de espécies, além da

presença de diversos táxons considerados endêmicos (Boldrini 2002). Boldrini (2009) aponta a

ocorrência de cerca de 2200 espécies campestres no Estado, distribuídas nos biomas Pampa e Mata

Atlântica. A alta diversidade biológica encontrada no Estado está, em grande parte, atrelada à

grande variabilidade geológica, topográfica, de pluviosidade, de temperatura e de disponibilidade de

água no solo (Boldrini 2009). Rambo (1954) chama a atenção para a riqueza e a importância

ecológica da flora campestre sul-brasileira, apontando que somente as espécies herbáceas de

Asteraceae presentes na região são mais numerosas do que toda a sua flora arbórea.

As formações campestres inseridas no bioma Mata Atlântica são caracterizadas por uma

distribuição em mosaico com a Floresta com Araucária e turfeiras, sendo que Andropogon lateralis

é a espécie de gramínea dominante na região, determinando sua fisionomia (Boldrini et al. 2009). A

pecuária extensiva, sobre campos manejados com queimadas no fim do inverno visando ao rebrote

da vegetação, é a atividade mais antiga da região, apesar de estar sendo gradativamente substituída

por lavouras e pelo plantio de espécies arbóreas exóticas (Boldrini et al. 2009). Há evidências de

que o uso do fogo como prática de manejo, devido à época e à frequência de aplicação, impeça o

desenvolvimento de espécies hibernais na região, diminuindo, assim, diversidade e sustentabilidade

da pastagem natural pelo predomínio de espécies entouceiradas de baixa qualidade forrageira

(Jacques 2003). Caporal & Eggers (2005), em levantamento da flora agrostológica de uma área de

Page 8: abordagem quali-quantitativa e funcional de vegetação campestre ...

9

campo natural preservado nos Campos de Cima da Serra do RS, apontam que 75% das espécies são

estivais e apenas 25% hibernais. Nabinger et al. (2000) apontam a prática da queimada como

agronomicamente improdutiva, tendo em vista o favorecimento de gramíneas “C4” e a conseqüente

diminuição de forragem durante o inverno. Llorens & Frank (2004) relacionam o uso do fogo no

inverno ou no início da primavera com diminuição da contribuição de espécies C3 e aumento das

C4. Boldrini et al. (2009) listam a ocorrência de 1161 táxons para os campos da região, sendo que o

maior número de espécies pertence à família Asteraceae (24%), seguida de Poaceae (20%),

Fabaceae e Cyperaceae (ambas com 7%). Do total de espécies, 107 são consideradas endêmicas, 76

estão inseridas na lista de espécies ameaçadas do RS e quatro eram novas para a ciência. Segundo

Boldrini (2009), o clima frio da região, aliado à alta pluviosidade, culminou na formação de

diversos endemismos de plantas herbáceas.

No Brasil, o bioma Pampa está restrito ao RS, apresentando continuidade florística com o

Uruguai e o Nordeste da Argentina (Burkart 1975; IBGE 2004; Boldrini 2009). Esta formação, que

cobre ca. 63% da superfície do Estado, é apontada por Burkart (1975) como uma das regiões mais

ricas em gramíneas do mundo. Os campos do bioma Pampa, apesar da aparente uniformidade

fisionômica, apresentam uma enorme diversidade de táxons e formações vegetacionais decorrente

da grande variabilidade edáfica da região (Boldrini 2009) e do seu caráter relictual (Behling 2002;

Bredenkamp et al. 2002). Boldrini (2009) divide os campos da região em sete tipologias, baseadas

em critérios florísticos, fisionômicos e edáficos: campos de barba-de-bode, de solos rasos, de solos

profundos, de areais, do centro do Estado, litorâneos e vegetação savanóide (ou campos da Serra do

Sudeste).

Apesar das conhecidas diferenças de solo, relevo, pluviosidade e composição florística

entre os campos inseridos nos dois diferentes biomas presentes no Estado, não há trabalhos com

abordagens quali-quantitativas estruturais comparando as duas formações.

Mesmo com o avanço do conhecimento em relação à flora campestre do Estado, estudos

quali-quantitativos que tenham como objetivo a caracterização dos diversos subtipos de campos que

Page 9: abordagem quali-quantitativa e funcional de vegetação campestre ...

10

existem no Estado (Boldrini 1997) são pontuais. Entre eles, Boldrini et al. (1998) realizaram um

levantamento fitossociológico da vegetação campestre presente em um morro de embasamento

granítico na região de Porto Alegre. Na mesma formação, Overbeck et al. (2006) avaliaram as

diferenças florísticas e sua relação com fatores abióticos, como propriedades de solo e distância da

borda da floresta. Levantamentos que empregaram metodologias semelhantes foram realizados por

Caetano (2003), Garcia (2005), Boldrini et al. (2008), e Ferreira & Setubal (2009), todos na Planície

Costeira do Estado. Caporal & Boldrini (2007) realizaram levantamento florístico e fitossociológico

em uma área de campo pastejada na Serra do Sudeste do RS. Freitas et al. (2009) realizaram

levantamento semelhante em um campo no sudoeste do Estado, sujeito à arenização. Estudos

comparativos envolvendo a estrutura das comunidades campestres dos biomas Pampa e Mata

Atlântica são imprescindíveis para futuras iniciativas de conservação e manejo dessas formações

naturais características da região. Todavia, estudos de longo prazo que avaliem variáveis estruturais

da comunidade atreladas à variável temporal são praticamente inexistentes. Overbeck et al. (2005)

avaliaram as mudanças entre dois anos em comunidades campestres após eventos de queimada,

utilizando uma área excluída de manejo para comparação. Os autores apontam que, na área

queimada, o turnover de espécies foi maior no primeiro ano, decaindo nos anos subseqüentes

devido ao aumento de cobertura de espécies cespitosas. Na área excluída de manejo, as espécies de

gramíneas cespitosas apresentaram maior dominância, e a área apresentou valores menores de

riqueza e diversidade quando comparada à área manejada com fogo.

Para entender os processos envolvidos na dinâmica desses ecossistemas campestres, tendo

em vista sua já conhecida complexidade, espécies e populações podem ser resumidas em padrões

funcionais gerais recorrentes (Walker 1992; Grime et al. 1996). Essa abordagem, embora tenha sido

introduzida na ciência há tempo (e.g. Raunkiaer 1934; Box 1981; ver revisão em Westoby 1998),

tem sido revisitada atualmente. Padrões consistentes de associação entre atributos de diferentes

plantas foram encontrados para floras locais (ver revisão em Diaz & Cabido 1997), e a perspectiva

de um método que permita classificar uma espécie, independentemente de sua distribuição, de

Page 10: abordagem quali-quantitativa e funcional de vegetação campestre ...

11

acordo com sua estratégia ecológica permitiria desvendar padrões mais gerais através da análise de

uma base de dados ampliada (Westoby 1998).

Partindo do princípio que certos atributos das plantas descrevam diferentes estratégias

ecológicas, esses atributos poderiam ser usados para generalizar mudanças na composição de

espécies ao longo do tempo (Weiher et al. 1999; Nygaard & Ejrnæs 2004). Mudanças significativas

de atributos funcionais ao longo de um período de sucessão já foram observadas em formações

campestres (Kahmen & Poschlod 2004; Lindborg & Eriksson 2005). Tendo em vista que a própria

existência atual dos campos com sua configuração atual está atrelada a processos sucessionais após

distúrbios como o fogo (Quadros & Pillar 2001, Behling 2002, Behling et al. 2004, Overbeck et al.

2005, Müller et al. 2007) e o pastejo (Senft et al. 1987; Coughenour 1991; Pillar & Quadros 1997),

estudos que enfoquem as mudanças florísticas e estruturais ao longo do tempo são imperativos.

No bioma Pampa, mais especificamente na região da Campanha, onde o manejo utilizando

fogo não é característico, o excesso de pressão de pastejo tem sido o problema (Boldrini 1997). Nos

campos da região, praticamente não há áreas excluídas de pastejo e, portanto, não há estudos que

enfoquem a dinâmica sucessional da vegetação campestre sem a presença do gado.

Como já abordado anteriormente, a fisionomia atual dos campos está relacionada a um

regime de distúrbios. Assim sendo, a exclusão de qualquer prática de manejo permite visualizar o

processo sucessional da comunidade, normalmente interrompido pelo manejo. Tendo em vista as já

conhecidas diferenças de manejo e flora entre as formações campestres dos biomas Pampa e Mata

Atlântica, espera-se que a resposta das comunidades de cada área seja diferente, de acordo com o

tipo (ou a ausência) de manejo aplicado. Esperam-se grupos de plantas que compartilham atributos

semelhantes entre si caracterizem a sucessão após cada distúrbio em cada área, e mesmo entre

diferentes períodos sucessionais.

Para estudar a estrutura e a dinâmica da vegetação campestre, em ambos os biomas,

levando em conta as práticas de manejo citadas e a variável temporal, é necessária a implantação de

Page 11: abordagem quali-quantitativa e funcional de vegetação campestre ...

12

parcelas permanentes, para que seja possível obter informações de longo prazo padronizadas e

futuramente comparáveis (Sanquetta 2008a).

Esta tese de doutorado foi realizada no âmbito do projeto de Pesquisa Ecológica de Longa

Duração (PELD) Campos Sulinos (CNPq 558282/2009-1). Este projeto teve início simultâneo ao

meu ingresso como aluno de doutorado no Programa de Pós-Graduação em Botânica da UFRGS.

No âmbito deste projeto, foram estabelecidos seis sítios de pesquisa em diferentes pontos do Estado

do Rio Grande do Sul (três no bioma Pampa e três no bioma Mata Atlântica), no ano de 2011. Na

Figura 1 são apresentados todos os sítios que atualmente compõem a rede PELD Campos Sulinos.

Esta tese foi realizada com base nos dados obtidos nos sitos 1, 2, 3, 4, 6 e 7 (Figura 1). Cada um

destes sítios constitui um bloco de amostragem, composto de três potreiros sob diferentes

tratamentos: manejo convencional, conservativo e exclusão de manejo (Figura 2). Detalhes sobre os

tipos de manejo são fornecidos nos capítulos subsequentes da tese. Desde 2011, todos os sítios vêm

sendo reamostrados anualmente.

Page 12: abordagem quali-quantitativa e funcional de vegetação campestre ...

13

Figura 1. Sítios que atualmente fazem parte da Rede PELD Campos Sulinos. Os sítios englobados

nesta tese são os do ‘experimento regional’, numerados de 1 a 7, com exceção do sítio número 5

que ainda não foi implementado.

Figura 2. Estrutura dos blocos de amostragem dos sítios da Rede PELD Campos Sulinos.

Page 13: abordagem quali-quantitativa e funcional de vegetação campestre ...

14

Os objetivos gerais desta tese, que são atingidos no conjunto dos artigos apresentados,

foram os seguintes: (i) caracterizar e comparar comunidades campestres inseridas nos biomas

Pampa e Mata Atlântica do Rio Grande do Sul em relação a sua composição florística e estrutural;

(ii) avaliar a aplicabilidade do conceito clássico de forma de vida sensu Raunkiaer como descritor

de padrões de vegetação campestre subtropical; (iii) identificar respostas das comunidades

campestres inseridas nos dois biomas a diferentes práticas de manejo com animais pastadores e à

exclusão de manejo e (iv) explorar as relações entre manejo, estrutura da vegetação, comunidades

de artrópodes e processos ecossistêmicos (decomposição).

A tese está estruturada em capítulos que correspondem a manuscritos de artigos, já

formatados para submissão em revistas científicas. O primeiro capítulo apresenta os dados obtidos

no primeiro ano de levantamento do projeto (2011), consistindo na descrição da vegetação dos

sítios, das diferenças entre sítios inseridos em diferentes biomas e em uma breve discussão sobre a

influência de variáveis edáficas nos padrões encontrados. Este manuscrito está nas normas da

revista Biodiveristy and Conservation. O segundo capítulo consiste em uma revisão do conceito de

formas de vida aplicado em plantas, seguido de uma proposta de categorias de formas de vida para

ser aplicada em ecossistemas campestres subtropicais. Este capítulo é finalizado com uma

comparação entre a eficácia desta classificação proposta e classificações clássicas e amplamente

utilizadas em descrever diferenças entre sítios com vegetação campestre sob níveis variáveis de

pastejo. O manuscrito está nas normas da revista Austral Ecology. No terceiro capítulo são

abordadas as relações entre manejo, estrutura da vegetação, comunidades de artrópodes e processos

ecossistêmicos. Este manuscrito está nas normas da revista Ecography. Por fim, o quarto capítulo

aborda aspectos da dinâmica da vegetação campestre sob diferentes manejos após quatro anos de

monitoramento. Este manuscrito está nas normas da revista Journal of Vegetation Science.

Page 14: abordagem quali-quantitativa e funcional de vegetação campestre ...

15

Referências

Adler PB, Milchunas DG, Laurenroth WK, Sala OE, Burke IC 2004. Functional traits of graminoids

in semi-arid steppes: a test of grazing histories. Journal of Applied Ecology 41: 653-663.

Behling H 2002. South and southeast Brazilian grassland during Late Quaternary times: a synthesis.

Palaegeography, Palaeclimatology, Palaeoecology 177:19-27.

Behling H, Pillar VD, Müller SC, Overbeck GE 2007. Late-Holocene fire history in a forest-

grassland mosaic in southern Brasil: Implications for conservation. Applied Vegetation Science

10:81-90.

Bertoletti JJ & Teixeira MB 1995. Centro de Pesquisas e Conservação da Natureza Pró-Mata.

Termo de Referência. Divulgações do Museu de Ciências e Tecnologia, UBEA/PUCRS 2:1-47.

Blanco CC, Sosinski Junior EE, Santos BRC, Abreu da Silva M, Pillar VD 2007. On the overlap

between effect and response plant functional types linked to grazing. Community Ecology 8:57-

65.

Boldrini II 1997. Campos do Rio Grande do Sul: caracterização fisionômica e problemática

ocupacional. Boletim do Instituto de Biociências UFRGS 56:1-39.

Boldrini II 2002. Campos sulinos: caracterização e biodiversidade. In: Araújo, EL Moura AN,

Sampaio EVSB, Gestinari LMS, Carneiro JMT (Eds.) Biodiversidade, conservação e uso

sustentável da flora Brasileira. Recife: Universidade Federal Rural de Pernambuco. p. 95-97.

Boldrini II 2009. A flora dos campos do Rio Grande do Sul. In: Pillar VDP, Müller SC, Castilhos

ZMC, Jacques AVA (Eds.) Campos Sulinos – conservação e uso sustentável da biodiversidade.

MMA. Brasília/DF.

Boldrini I.I. & Eggers L. 1996. Vegetação campestre do sul do Brasil: resposta e dinâmica de

espécies à exclusão. Acta Bot. Bras. 10: 37-50.

Page 15: abordagem quali-quantitativa e funcional de vegetação campestre ...

16

Boldrini II, Eggers L, Mentz LA, Mioto STF, Matzenbacher NI, Longhi-Wagner HM, Trevisan R,

Schneider AA, Setúbal RB 2009. Flora. In: Boldrini II 2009 (Org.). Biodiversidade dos campos

do Planalto das Araucárias. MMA. Brasília/DF.

Boldrini II, Trevisan R, Schneider AA 2008. Estudo florístico e fitossociológico de uma área às

margens da lagoa do Armazém, Osório, Rio Grande do Sul, Brasil. Revista Brasileira de

Biociências 6(4): 355-367.

Box EO 1981. Macroclimate and plant forms: an introduction to predictive modelling in

phytogeography. Junk, The Hague.

Braun-Blanquet J 1979. Fitosociologia: bases para el estudio de las comunidades vegetales.

Madrid: H. Blume Ediciones. 820 p.

Bredenkamp GJ, Spada F, Kazmierczak E 2002. On the origin of northern and southern hemisphere

grasslands. Plant Ecology 16:209-229.

Bullock JM 1996. Plant competition and population dynamics. In: The Ecology and Management of

Grazing Systems (eds. Hodgson J & Illius AW). CAB International Wallingford, pp. 69-100.

Burkart A 1975. Evolution of grasses and grasslands in South America. Taxon 24: 53-66.

Berlato MA 1970. Análise de alguns elementos componentes do agroclima do Estado do Rio

Grande do Sul. Turralba: IICA. 117p. Tese (Mestrado em climatologia). IICA.

Berlato MA, Fontana DC, Puchalski L 2000. Precipitação pluvial normal e riscos de ocorrência de

dificiência pluviométrica e deficiência hidrica no Rio Grande do Sul: Ênfase para a metade sul

do Estado. In: Flávio Gilberto Hertel. (Org.). Seminário sobre água na produção de frutíferas. 68

ed. Pelotas-RS: Embrapa Clima Temperado, 2000, v. 68, p. 67-81.

Cabrera AL & Willink A 1980. Biogeografia da America Latina. 2 ed. OEA, Washington, 117 p.

Caporal, FJM & Eggers, L. 2005. Poaceae na área do Centro de Pesquisas e Conservação da

Natureza Pró-Mata, São Francisco de Paula, Rio Grande do Sul, Brasil. Iheringia 60: 141-150.

Caporal JFM & Boldrini II 2007. Florística e fitossociologia de um campo manejado na

Serra do Sudeste, Rio Grande do Sul. Rev. Bras.Bioci. 5:37-44.

Page 16: abordagem quali-quantitativa e funcional de vegetação campestre ...

17

Caetano VL 2003. Dinâmica sazonal e ftossociologia da vegetação herbácea de uma baixada úmida

entre dunas, Palmares do Sul, Rio Grande do Sul, Brasil. Iheringia, sér. Bot., 58(1): 81-102.

Cervi AC, Linsingen L, Hatschbach G, Ribas OS 2007. A vegetação do Parque Estadual de Vila

Velha, Municipio de Ponta Grossa, Paraná, Brasil. Boletim do Museu Botânico Municipal 69:1-

52.

Chao A 1984. Nonparametric estimation of the numbers of classes in a population. Scandinavian

Journal of Statistics 11:265-270.

Coughenour MB 1991. Spatial components of plant-herbivore interactions in pastoral, ranching, and

native ungulate ecosystems. Journal Range Management 44: 530-541.

Cordeiro & Hasenack 2009. Cobertura vegetal atual do Rio Grande do Sul. In: Pillar VDP, Müller

SC, Castilhos ZMC, Jacques AVA (Eds.) Campos Sulinos – conservação e uso sustentável da

biodiversidade. MMA. Brasília/DF.

Crawshaw D, Dall’Agnol M, Cordeiro JLP, Hasenack H 2007. Caracterização dos campos Sul-Rio-

Grandenses: uma perspectiva da Ecologia da Paisagem. Boletim Gaúcho de Geografia 33: 233-

252.

Dahlgren JP, Eriksson O, Bolmgren K, Strindell M, Ehrlén J 2006. Specific leaf area as a superior

predictor of changes in field layer abundance during forest succession. J Veg Sci 17: 577-582.

Díaz S, Acosta A, Cabido M 1992. Morphological analysis of herbaceous communitites under

different grazing regimes. Journal of Vegetation Science 3: 689-696.

Díaz S, Cabido M 1997. Plant functional types and ecosystem function in relation to global change.

J Veg Sci 8:463–474

Díaz S, Cabido M, Zak M, Carretero EM, Araníbar J 1999. Plant functional traits, ecosystem

structure and land-use history along a climatic gradient in central-western Argentina. Journal of

Vegetation Science 10: 651-660.

Page 17: abordagem quali-quantitativa e funcional de vegetação campestre ...

18

Durigan, G 2003. Métodos para análise de vegetação arbórea. In: Cullen-JR L, Pádua CV, Rudran R

(Org). Métodos de estudos em biologia da conservação & manejo da vida silvestre. Curitiba: Ed.

da UFPR. p. 455-479.

Ferreira PMA & Setúbal RB 2009. Florística e fitossociologia de um campo natural no município

de Santo Antonio da Patrulha, Rio Grande do Sul, Brasil. Rev. Bras. Bioci. 7(2):195-204.

Freitas EM, Boldrini II, Müller SC, Verdum R 2009. Florística e fitossociologia da vegetação de um

campo sujeito à arenização no sudoeste do Estado do Rio Grande do Sul, Brasil. Acta bot. bras.

23(2): 414-426.

Garcia EN 2005. Subsídios à conservação de campos no norte da Planície Costeira do Rio Grande

do Sul, Brasil. 110 f. Tese de Doutorado.Universidade Federal do Rio Grande do Sul, Porto

Alegre. 2005.

Garnier E, Cortez J, Billès G, Navas M, Roumet C, Debussche M, Laurent G, Blanchard A, Aubry

D, Bellmann A, Neill C, Toussaint J 2004. Plant functional markers capture ecosystem properties

during secondary succession. Ecology 85(9):2630-2637.

Gomes JF, Longhi SJ, Araujo MM, Brena DA 2008. Classificação e crescimento de unidades de

vegetação em Floresta Ombrófila Mista, São Francisco de Paula, RS. Ciência Florestal 18:93-

107.

Grime JP, Hodgson JG, Hunt R, Thompson K, Hendry GAF, Campbell BD, Jalili A, Hillier SH,

Díaz S, Burke MJW 1996. Functional types: Testing the concept in Northern England. In:

Smith,T.M., Shugart, H.H. & Woodward, F.I. (eds.) Plant functional types, pp. 123- 131.

Cambridge University Press, Cambridge.

IBGE (Instituto Brasileiro de Geografia e Estatística) 2006. Censo agropecuário 1995-1996. IBGE.

(acessado em setembro de 2009).

IBGE (Instituto Brasileiro de Geografia e Estatística). 2004. Mapa da vegetação do Brasil e Mapa

de Biomas do Brasil. URL www.ibge.gov.br

Page 18: abordagem quali-quantitativa e funcional de vegetação campestre ...

19

Jacques AVA 2003. A queima das pastagens naturais – efeitos sobre o solo e a vegetação. Ciência

Rural 33: 177-181.

Jüngblut M & Pinto LFS 1997. Levantamento de solos do Centro de Pesquisa e Conservação da

Natureza Pró- Mata. Divulgação do Museu de Ciências e Tecnologia, UBEA/PUCRS 3: 29-94.

Kahmen S & Poschlod P 2004. Plant functional trait responses to grassland succession over 25

years. J. Veg. Sci. 15: 21-32.

Kent M & Coker P 1995. Vegetation description and analysis: a practical approach. Chichester:

John Wiley. 363 p.

Kozera C 2008. Florística e fitossociologia de uma Formação Pioneira com Influência Fluvial e de

uma Estepe Gramíneo-Lenhosa em diferentes unidades geopedológicas, município de Balsa

Nova, Paraná – Brasil. Tese (Doutorado em Engenharia Florestal). Setor de Ciências Agrárias,

Universidade Federal do Paraná, 267p.

Landsberg J, Lavorel S, Stol J 1999. Grazing response among undestorey plants in arid rangelands.

Journal of Vegetation Science 10: 683-696.

Lavorel S, McIntyre S, Grigulis K 1999. Plant response to disturbance in a Mediterranean

grassland: how many functional groups? Journal of Vegetation Science 10: 661-672.

Leivas JF, Berlato M, Fontana D 2006. Risco de deficiência hídrica decendial na metade sul do

Estado do Rio Grande do Sul. Revista Brasileira de Engenharia Agrícola e Ambiental (Online)

10:397-407.

Lindborg R & Eriksson O 2005. Functional response to land use change in grasslands comparing

species and trait data. Ecoscience 12: 183-191.

Llorens, E.M. & Frank, E.O. 2004. El fuego en la provincia de La Pampa. In: Kunst, C., Bravo, S.

& Panigatti, J.L. (eds.) Fuego en los ecosistemas argentinos, pp. 259-268. Instituto Nacional de

Tecnología Agropecuaria, Santiago del Estero.

Page 19: abordagem quali-quantitativa e funcional de vegetação campestre ...

20

Machado RE 2004. Padrões vegetacionais em capões de Floresta com Araucária no Planalto

Nordeste do Rio Grande do Sul, Brasil. Dissertação de Mestrado, Programa de Pós-graduação em

Ecologia, Universidade Federal do Rio Grande do Sul, Porto Alegre, p. 164.

Maraschin GE 2001. Production potential of South America grasslands. In: International Grassland

Congress São Paulo, pp. 5-15.

Mueller-Dombois D & Ellenberg H 1974. Aims and methods of vegetation ecology. New York:

John Wiley. 547p.

Müller SC, Overbeck GE, Pfadenhauer J, Pillar VD 2007. Plant functional types of woody species

related to fire disturbance in forest-grassland ecotones. Plant Ecology 189:1-14.

Nabinger C, Moraes A, Maraschin GE 2000. Campos in Southern Brazil. In: Grassland

ecophysiology and grazing ecology (eds. Lemaire G, Hodgson JG, Moraes A & Maraschin GE).

CABI Publishing Wallingford, pp. 355-376.

Noble IR & Gitay H 1996. A functional classification for predicting the dynamics of landscapes. J.

Veg. Sci. 7: 329-336.

Narvaes IS, Longhi SJ, Brena DA 2008. Florística e classificação da regeneração natural em

Floresta Ombrófila Mista na Floresta Nacional de São Francisco de Paula, RS. Ciência Florestal

18:233-245.

Nygaard B & Ejrnæs R 2004. A new approach to functional interpretation of vegetation data. J.

Veg. Sci. 15: 49-56.

Oliveira JM & Pillar VD 2004. Vegetation dynamics on mosaics of Campos and Araucaria forest

between 1974 and 1999 in Southern Brazil. Community Ecology 5: 197-202.

Overbeck GE, Müller SC, Pillar VD, Pfadenhauer J 2005. Fine-scale post-fire dynamics in southern

Brazilian subtropical grassland. Journal of Vegetation Science 16:655-664.

Overbeck GE, Müller SC, Pillar VD, Pfadenhauer J 2006. Floristic composition, environmental

variation and species distribution patterns in burned grassland in southern Brazil. Brazilian

Journal Biology 66(4):1073-1090.

Page 20: abordagem quali-quantitativa e funcional de vegetação campestre ...

21

Overbeck GE, Müller SC, Fidelis A, Pfadenhauer J, Pillar VD, Blanco CC, Boldrini II, Both R,

Forneck ED 2007. Brazil’s neglected biome: The South brazilian Campos. Perspect. Plant Ecol.

Evol. Systematics 9:101-116.

Pillar VD 2006. MULTIV sofware para análise multivariada, testes de aleatorização e

autoreamostragem “bootstrap", v. 2.4.2. Porto Alegre: Departamento de Ecologia, UFRGS

Pillar VD & Quadros FLF 1997. Grassland-forest boundaries in southern Brazil. Coenoses 12: 119-

126.

Pillar VD 1999a. How sharp are the classifications? Ecology 80(8):2508-2516.

Pillar VD 1999b. The bootstrap ordination re-examined. Journal of Vegetation Science 10:895-905.

Pillar VD & Orlóci L 1993. Character-Based Community Analysis; the Theory and an Application

Program. The Hague: SPB Academic Publishing

Pillar VD & Orlóci L 1996. On randomization testing in vegetation science:multifactor comparisons

of relevé groups. Journal of Vegetation Science 7:582-592.

Pillar VD & Sosinski EE Jr 2003. An improved method for searching plant functional types by

numerical analysis. J Veg Sci 14:323–332.

Pillar VD 2004. SYNCSA software for character-based community analysis, v. 2.2.4. -Departamento

de Ecologia, UFRGS

Quadros FLF & Pillar VD 2001. Dinâmica vegetacional em pastagem natural submetida a

tratamentos de queima e pastejo. Ciência Rural 31(5):863-868.

Quadros FLF, Trindade JPJ, Borba M 2009. A abordagem funcional da ecologia campestre como

instrumento de pesquisa e apropriação do conhecimento pelos produtores rurais. In.: Pillar VDP,

Müller SC, Castilhos ZMC, Jacques AVA (Eds.) Campos Sulinos – conservação e uso sustentável

da biodiversidade. MMA. Brasília/DF.

Rambo B 1942. A fisionomia do Rio Grande do Sul: ensaio de monografia natural. Porto Alegre,

Oficina Gráfica da Imprensa Oficial.

Rambo B. 1954. Análise histórica da flora de Porto Alegre. Sellowia 6:9-111.

Page 21: abordagem quali-quantitativa e funcional de vegetação campestre ...

22

Raunkiaer C 1934. The life forms of plants and statistical plant geography; being the collected

papers of C. Raunkiaer. Clarendon Press, Oxford.

REDEMAP - Rede de Parcelas Permanente dos Biomas Mata Atlântica e Pampa.

www.redemap.org. Acessado em 29/09/2009.

Rodríguez C, Leoni E, Lezama F, Altesor A 2003. Temporal trends in species composition and

plant traits in natural grasslands of Uruguay. Journal of Vegetation Science 14: 433-440.

Sanquetta CR 2008a. RedeMAP - Manual para instalação e medição de parcelas permanentes nos

biomas Mata Atlântica e Pampa. Curitiba. 44p.

Sanquetta CR 2008b (ed.). Experiências de monitoramento no bioma mata atlântica com uso de

parcelas permanentes. Curitiba. 338p.

Senft RL, Coughenour MB, Bailey DW, Rittenhouse LR, Sala OE, & Swift DM 1987. Large

herbivore foraging and ecological hierarchies. BioScience 37:789-799.

SisPP – Sistema Nacional de Parcelas Permanentes.

http://www.cnpf.embrapa.br/pesquisa/sispp/SisPP.htm. Acessado em 29/09/2009.

Sosinski Jr EE & VD Pillar 2004. Respostas de tipos funcionais à intensidade de pastejo em

vegetação campestre. Pesquisa Agropecuária Brasileira 39: 1-9.

Suttie JM, Reynolds SG, Batello C 2005. Grasslands of the World. FAO, Rome.

Teixeira MB, Coura-Neto AB, Pastore U, Rangel Filho ALR 1986. Vegetação. In: Levantamento de

recursos naturais (ed. IBGE). IBGE Rio de Janeiro, pp. 541-632.

Walker BH 1992. Biodiversity and ecological redundancy. Conserv. Biol. 6: 18-23.

Weiher E, van der Werf A, Thompson K, Roderick M, Garnier E, Eriksson O 1999. Challenging

Theophrastus: a common core list of plant traits for functional ecology. J Veg Sci 10:609–620

Westoby M 1998. A leaf-height-seed (LHS) plant ecology strategy scheme. Plant and Soil 199:

213-227.

Whittaker RH 1975. Communities and ecosystems, 2nd edn. Macmillan Publishing CO. Inc. : New

York.

Page 22: abordagem quali-quantitativa e funcional de vegetação campestre ...

23

Page 23: abordagem quali-quantitativa e funcional de vegetação campestre ...

24

Differences in grassland vegetation from two South Brazilian biomes and implications for

conservation

Pedro M.A. Ferreira, Bianca O. Andrade, Gerhard E. Overbeck, Ilsi I. Boldrini

P.M.A. Ferreira (corresponding author), B.O. Andrade, G.E. Overbeck, I.I. Boldrini

Universidade Federal do Rio Grande do Sul, Programa de Pós Graduação em Botânica, Av. Bento

Gonçalves 9500 Bloco IV, P. 43432, CEP 91501-970, Porto Alegre, RS, Brazil.

e-mail: [email protected]

phone: +55 51 3308 7555

Page 24: abordagem quali-quantitativa e funcional de vegetação campestre ...

25

Abstract 1

Conservation is a global concern, and can produce more effective results when encompassing 2

simultaneous conservation and use of biodiversity. This paradigm highlights the importance of 3

natural grasslands for conservation, since these ecosystems may be examples of sustainability by 4

allying profitable use and maintenance of biodiversity. Grasslands in Southern Brazil are inserted in 5

two biomes: Pampa and Atlantic Forest. Quantitative studies focusing on floristic and structural 6

differences between grasslands in both biomes are so far lacking. This paper aims to evaluate 7

differences in plant composition, richness, diversity and structural variables between natural 8

grasslands from Pampa and Atlantic Forest biomes, and the implications of these differences for 9

conservation. We also evaluated the correlation between soil features and vegetation patterns. Data 10

were collected in 162 sampling units distributed in nine paddocks in six sites, three per biome. Data 11

were submitted to cluster and ordination analyses. Relationships between soil features and 12

vegetation were assessed with linear regression using ordination axes. Relationships between 13

structural variables and plant community were estimated using correlation analysis. Sampling 14

resulted in 382 plant taxa from 40 families (ca. 17% of the regional grassland flora). Sites between 15

biomes shared 28 families and only 15% species. Average richness and diversity were higher in 16

Pampa sites. Cluster and ordination analyses revealed two sharp groups among sampling units, 17

consistent with biome separation. Dominance was higher in Atlantic Forest sites than in Pampa 18

sites. We inferred that this vegetation structure is the result of past and present differences in 19

management, soil and climate. The implication of our results is that conservation efforts must be 20

equally focused on grasslands from both biomes in order to target: (1) a representative set of 21

species, (2) different vegetation structures and (3) potentially different ecological processes and 22

services. Also, we suggest that management planning that includes grazing and/or fire must be 23

mandatory in Brazilian conservation units encompassing grasslands. 24

25

Keywords: management, soil-plant relationships, conservation units, sustainable use, grassland 26

biodiversity, disturbance regime. 27

Page 25: abordagem quali-quantitativa e funcional de vegetação campestre ...

26

Introduction 1

2

Biodiversity conservation has been repeatedly pointed out as a priority task of societies 3

worldwide (Sachs et al. 2009; Larigauderie and Mooney 2010). Although it was originally focused 4

on preserving species, conservation is today a broad multidisciplinary science, and well-planned 5

conservation efforts encompass species, landscapes and ecological processes alike (Heywood and 6

Iriondo 2003). Human actions during the past few centuries have significantly reduced natural 7

ecosystems in area and services offered (Millennium Ecosystem Assessment 2005), emphasizing 8

the importance of conservation. Moreover, scientists and political actors are slowly converging 9

towards the idea of allying conservation and use of natural resources (Aronson et al. 2006). 10

Therefore, ecosystems dominated by natural grassland landscapes may be key assets for 11

conservation, since they are natural forage sources for domestic herbivores (Hodgson 1990; 12

Nabinger et al. 2000; Bilenca and Miñarro 2004) and their biodiversity is not impaired by this 13

disturbance, at least when it does not occur in extreme intensities. In fact, grassland biodiversity and 14

conservation status may be positively influenced by grazing (Olff and Ritchie 1998; Sebastià et al. 15

2008), thus configuring a ‘natural’ example of sustainable use. 16

Grassland ecosystems cover large areas in Southern South America. The Río de la Plata 17

Grasslands extend over ca. 750,000 km2 in Argentina, Uruguay and Southern Brazil (Soriano et al. 18

1992; Bilenca and Miñarro 2004). Within Brazilian territory, these grasslands determine the 19

landscape of the southern half of Rio Grande do Sul, the southernmost Brazilian state, in the region 20

defined locally as the Pampa biome (IBGE 2004). However, grasslands are also present in the 21

Atlantic Forest biome, in higher altitudes, milder climates and shaping mosaics with forests 22

(Overbeck et al. 2007). The contact zone between these biomes is located near the 30oS parallel, a 23

known threshold between tropical and subtropical/temperate vegetation (Cabrera and Willink 24

1980). This region is also a transition point considering geomorphology and soil types. Soils in the 25

Atlantic Forest biome are mostly derived from basalt and show low pH values on average, whereas 26

Page 26: abordagem quali-quantitativa e funcional de vegetação campestre ...

27

there is a higher diversity of soil types in the Pampa biome (Streck et al. 2008). Soils have been 1

long hypothesized to influence vegetation patterns in South Brazilian grasslands, as it has been 2

found for Uruguayan grasslands by Lezama et al. (2006), although no study has specifically focused 3

on that relationship so far. 4

Temperate grasslands and savannas figure among the world’s most critically endangered 5

ecosystems, with 45.8% rate of conversion and only 4.6% of protection (Hoekstra et al. 2005). 6

Conservation of South American ecosystems has long been identified as a problem (Mares 1986), 7

and still remains overlooked (e.g., Ramirez-Villegas et al. 2012). Conservation of South Brazilian 8

grasslands (locally known as campos) has been neglected. Less than 0.5% of the South Brazilian 9

grasslands are protected in conservation units, and most of them are inserted in the Atlantic Forest 10

biome (Overbeck et al. 2007). Moreover, there are no conservation units under IUCN’s categories I 11

to IV (Olson and Dinerstein 1998) in grasslands within the Pampa biome, increasing their 12

vulnerability to land conversion and suppression of natural vegetation. There was a decrease of ca. 13

25% in total natural grasslands in Southern Brazil between 1970 and 2000 (Nabinger et al. 2000). 14

Today, only 50% of original grassland cover in Rio Grande do Sul remains, and land conversion for 15

human use is estimated to be of 1,000 km2 per year (Cordeiro and Hasenack 2009). Grasslands in 16

the region have evolved under different levels of grazing (Milchunas et al. 1988) and fire 17

disturbances (Behling et al. 2004; Behling et al. 2005), have been used since the seventeenth 18

century as forage source for cattle breeding (Pillar and Quadros 1997) and are still used as such 19

today (Nabinger et al. 2000). The use of natural grasslands as forage sources for extensive livestock 20

breeding maintains grassland diversity, providing that adequate stocking rates are used (Hodgson 21

1990; Nabinger et al. 2009). In fact, conservation of grassland ecosystems around the world often 22

involves herbivory and/or fire as management tools (e.g., Fuhlendorf et al. 2006; Hampicke and 23

Plachter 2010; Houston 1982; Meagher 1973). However, land management within Brazilian 24

conservation units is still a taboo subject, mostly due to misguided ecological concepts (Pillar and 25

Vélez 2010) – even though, in fact, the very existence of these grassland ecosystems is linked to 26

Page 27: abordagem quali-quantitativa e funcional de vegetação campestre ...

28

disturbances regimes such as grazing and fire (Behling et al. 2004; Quadros and Pillar 2001). Land 1

management of grazed grasslands differs between biomes in Southern Brazil. In the Pampa, 2

stocking rates are usually higher, fire is seldom used as a managing tool and mowing is used to 3

control undesirable species. In the Atlantic Forest, stocking rates are usually lower, and undesirable 4

species and accumulated dry biomass are controlled with yearly fire (Maraschin 2001; Nabinger et 5

al. 2000). There is evidence that the use of fire with such periodicity reduces diversity and overall 6

foraging value by removing intolerant species and favoring tussock C4 species (reviewed by 7

Jacques 2003). It is likely that the difference in historical management between biomes has impacts 8

on present vegetation structure, although this question is yet to be directly addressed. 9

Plant species diversity in South Brazilian grasslands is extremely high. Boldrini (1997) 10

estimated 2,200 grassland plant species for Rio Grande do Sul state alone, and ongoing work by the 11

same author and colleagues will soon provide an updated species list with roughly 2.600 taxa 12

(Boldrini et al. unpubl.). Although the identity of plant species present in these ecosystems is 13

relatively well known (Boldrini 2009), possible structural and floristic differences between 14

grasslands inserted in the Pampa and in the Atlantic Forest biomes have not been elucidated. 15

Evaluating such differences will provide valuable tools for future conservation efforts, since these 16

must consider the representativeness of grassland ecosystems from different biomes (and possibly 17

different regions within each biome), different management types and different disturbance history. 18

Moreover, if floristics and/or community structure differs between grasslands from each biome, 19

conservation efforts should encompass different strategies in each biome in order to maximize 20

conservation of biodiversity, landscapes and ecological processes. 21

The objective of this paper is to evaluate differences between natural grassland areas from 22

the Pampa and the Atlantic Forest biomes, and the implications of these differences for 23

conservation. We hypothesize that these formations differ regarding plant composition, richness, 24

diversity and structural variables. We also briefly consider the influence of soil features over 25

vegetation patterns. 26

Page 28: abordagem quali-quantitativa e funcional de vegetação campestre ...

29

1

Material and Methods 2

3

Data collection took place in late 2010 and early 2011, in natural grasslands in Rio Grande 4

do Sul, Southern Brazil. Vegetation surveys were carried out at six sites, three in the Pampa biome 5

and three in the Atlantic Forest biome. Sites in the Pampa are inserted in farms predominantly 6

focused on cattle breeding. Sites in the Atlantic Forest are inserted in Conservation Units. 7

Acronyms for each sampling site were given according to municipalities or Conservation Units they 8

were inserted. Pampa sites were Aceguá (ACE; 31o38’55”S, 54o09’26”W), Alegrete (ALE; 9

30o04’08”S, 55o59’27”W) and Lavras do Sul (LAV; 30o41’55”S, 53o58’11”W). Atlantic Forest 10

sites were Aparados da Serra National Park (APA; 29o08’10”S, 50o09’21”W), Aratinga Ecological 11

Station (ARA; 29o23’31”S, 50o14’30”W) and Tainhas State Park (TAI; 29o05’40”S, 50o22’03”W). 12

Mean altitude in Pampa sites was 215 m AMSL and 930 m in Atlantic Forest sites. Grasslands in all 13

sites are under cattle grazing, and have been under grazing for many years, with no known record of 14

land conversion. In Atlantic Forest sites, yearly fires have also been used for many years, following 15

a widespread regional management technique. Although grazing intensity (livestock units per area) 16

varies from site to site, overall it is lighter in Atlantic Forest sites. We estimated grazing pressure in 17

each site in animal units (AU = 450 kg of live weight) per hectare. Grazing pressure in each site 18

was: ACE = 1.05 AU/ha, ALE = 0.9, LAV = 0.85, APA = 0.9, ARA = 0.6 and TAI = 0.45). 19

At each site we sampled three paddocks of 0.5 ha each using nine 1m2 permanent plots 20

(systematically allocated in the paddock in a 3x3 grid with 17 m between plots), summing up to 162 21

sampling units across the six sites. This sampling layout was designed for an ongoing long-term 22

ecological research (PELD Campos Sulinos; CNPq 558282/2009-1), and this paper reports results 23

from the first season of sampling (Southern hemisphere summer 2010/2011). In each sampling unit, 24

we surveyed all plant species that were present and estimated their cover using the decimal scale of 25

Londo (1976). We also estimated cover of bare soil, litter, rock outcrops and overall vegetation 26

Page 29: abordagem quali-quantitativa e funcional de vegetação campestre ...

30

cover per sampling unit. We calculated relative frequency and cover for each species (Ellenberg and 1

Mueller-Dombois 1974). 2

We submitted vegetation data to cluster analysis with sum of squares as clustering criterion 3

(Orloci 1967) and Principal Coordinate Analysis using chord distance as dissimilarity measure 4

between sampling units. We used mean values of species cover in sampling units per paddock in all 5

analyses. We performed additional multivariate analyses using presence/absence data to assess the 6

importance of species composition alone over vegetation patterns. We evaluated the presence of 7

sharp groups within cluster groups and stability and significance of ordination axes with bootstrap 8

resampling methods (Pillar 1998; Pillar 1999a, b). For comparisons of diversity between sites we 9

calculated the Shannon diversity index and evenness (Magurran 1988), and also used Hill’s 10

diversity profiles (Hill 1973; Tóthmérész 1995). 11

We collected soil samples in each sampling unit up to 10 cm depth. Air-dried soil samples 12

were used for chemical and textural analysis (Silva 1999). The pH value was determined in water 13

solution (1:1). Exchangeable cations, Mg+2, Ca+2 and Al+3 were extracted with KCl 1molL-1. P and 14

K were determined using the Mehlich I extraction method. Cation exchange capacity (CEC) was 15

determined at pH 7. Organic carbon content of the soil was measured using the wet combustion 16

method. Clay content was determined by densimeter method. In this paper, we do not wish to 17

discuss soil characteristics in detail, since in-depth soil analysis will be discussed in a separate study 18

(Andrade et al. unpublished). Rather, we used summarized soil data to estimate the influence of 19

abiotic factors on grassland vegetation parameters. To do so, we used the ordination scores of 20

sampling units from the first ordination axis obtained in a Principal Coordinate Analysis of 21

sampling units described by an abiotic matrix containing the variables described above (162 22

sampling units described by 17 soil variables, see ESM 1). Then we explored the relationships 23

between this axis and the axis obtained in the ordination of vegetation data (162 sampling units 24

described by abundances of 382 plant species). Both vectors were normalized, submitted to 25

Page 30: abordagem quali-quantitativa e funcional de vegetação campestre ...

31

correlation analysis (Pearson product-moment) and fitted with a linear regression model using the 1

vegetation axis as dependent variable. 2

Differences between sites regarding richness, diversity, species composition and cover and 3

structural variables were tested using randomization tests with 10,000 bootstrap resampling 4

iterations (Pillar and Orlóci 1996). For comparisons between species composition and cover we 5

used raw matrices (sampling units described by species mean percentage cover values). Correlation 6

between variables was evaluated with Pearson product-moment correlation coefficient, using 7

permutation tests to assess statistical significance. Only significant correlations were presented and 8

discussed (P<0.05). Prior to correlation analysis, data subsets were submitted to Shapiro-Wilk’s 9

normality test, and vectors that failed the test were normalized. Analyses were conducted with the 10

software Multiv (Pillar 1997) and on the R platform (R Development Core Team 2012). 11

12

Results 13

14

We found 382 plant taxa distributed in 40 families among the six sampling sites. Families 15

with highest overall species richness and average cover were Poaceae, Asteraceae and Cyperaceae 16

in both biomes, although the richness/cover ratio varied between biomes (Figure 1). The two groups 17

of sites from each biome shared 28 families and had eight exclusive families each. Both sets of 18

three sites within biomes had 27 species in common, whereas the two groups of sites from each 19

biome shared 57 species overall (Figure 2). 20

Average species richness and diversity were overall higher in the Pampa biome, both at 21

sampling unit and paddock levels. Richness values showed similar patterns in both biomes, with 22

one site encompassing more species and two sites with less species and no significant difference 23

between each other (Table 1). Considering only the Shannon index, diversity was similar among 24

Atlantic Forest sites, whereas it was different among Pampa sites (Table 1). However, the diversity 25

profiles revealed that diversities between biomes are different at low alpha values. Diversities were 26

Page 31: abordagem quali-quantitativa e funcional de vegetação campestre ...

32

equivalent within the Pampa biome, whereas within the Atlantic Forest they differed with 1

increasing alpha values (Figure 3). Randomization tests comparing species composition and cover 2

resulted in significant differences between biomes at the paddock and site levels (P<0.05). All 3

pairwise comparisons between sites of the same biome were not significant. 4

Differences between biomes were also consistent regarding structural variables. Vegetation 5

cover and height showed higher values in site from the Atlantic forest biome, whereas open soil 6

showed the opposite pattern (Table 2). Moreover, vegetation cover was more uniform in Atlantic 7

Forest sites in comparison with Pampa sites. Cover of litter and rock outcrops did not differ 8

between biomes, and showed significant differences only at the TAI site among Atlantic Forest sites 9

(Table 2). Atlantic Forest sites showed high dominance of Andropogon lateralis, an erect tussock 10

grass, whereas the prostrate grass Paspalum notatum was the most representative at Pampa sites 11

(Table 3). 12

Across sites and biomes, mean bare soil was positively correlated to mean plant species 13

richness (r=0.75, P=0.0013), Shannon diversity (r=0.93, P<0.001) and evenness (r=0.77, P<0.001). 14

Evenness was negatively correlated to mean vegetation cover (r=-0.63, P<0.01) and positively 15

correlated to rock outcrops (r=0.61, P<0.01). Grazing pressure was correlated to vegetation height 16

(r=0.89, P<0.01), Shannon diversity (r=0.92, P<0.01) and species richness (r=0.68, P<0.01). 17

Cluster analysis revealed two sharp groups among paddocks. These groups are consistent 18

with the biome separation. However, the distance between paddocks was different within each 19

biome. Dissimilarity values between Pampa paddocks are on average twice as large as between 20

Atlantic Forest paddocks, which resulted in a perfect clustering match for sites from the former and 21

not for the latter (data not shown). Principal Coordinate Analysis revealed a pattern not entirely 22

consistent with biome separation. Although Atlantic Forest paddocks clustered in the left side of the 23

scatterplot, Pampa sites clustered in two groups: one comprising ACE and LAV paddocks, and the 24

other comprising ALE paddocks (Figure 4). Ordination and cluster analyses performed with 25

presence/absence data resulted in a similar pattern, although biome separation was clearer. 26

Page 32: abordagem quali-quantitativa e funcional de vegetação campestre ...

33

The first standardized ordination axis containing vegetation information was highly and 1

significantly correlated to the first ordination axis containing soil information (r=0.86; see methods 2

for details on soil data). The linear model using the vegetation axis as dependent variable showed 3

that vegetation composition and structure can be predicted by soil features across our sampling units 4

(Figure 5). Also, vegetation structural parameters were strongly correlated to individual soil 5

variables, but these relationships will be discussed elsewhere (Andrade et al. unpubl.). 6

7

Discussion 8

9

We aimed at revealing principal differences in biotic composition of south Brazilian 10

grassland sites included in two different biomes, Pampa and Atlantic Forest, for the first time based 11

on analyses of quantitative data. Our results indicate that grasslands in the Pampa biome differ from 12

those in the Atlantic Forest considering plant species composition and cover, richness (Table 1) and 13

vegetation structural variables (Table 2). Cluster and ordination analyses (Figure 4) also showed a 14

clear distinction between paddocks/sites from different biomes. The high dominance of Andropogon 15

lateralis in Atlantic Forest sites (Table 3) corroborates earlier descriptive studies of the local flora 16

(Boldrini 1997, 2009; Boldrini et al. 2009; Boldrini and Longhi-Wagner 2011), and partly explains 17

the lower diversity values and structural differences between sites. This species forms dense 18

tussocks, defining the landscape of this region. Dominance in Pampa sites, on the other hand, is 19

more diluted among species (also shown in Figure 3, with similar diversity with increasing alpha 20

values; see discussion on diversity profiles below), and the dominant grass is the prostrate 21

Paspalum notatum, which also corroborates previous descriptive studies (Boldrini 1997; Díaz et al. 22

1992; Pinto et al. 2013; Rodríguez et al. 2003). It is important to mention that the Atlantic Forest 23

Biome extends northwards along the entire Brazilian coast, whereas our sampling was restricted to 24

grasslands in the southern portion of the biome. Grasslands in the northern Atlantic Forest, 25

however, are usually related to high elevations, inserted in a more continuous forest matrix, have 26

Page 33: abordagem quali-quantitativa e funcional de vegetação campestre ...

34

different structure and floristic composition and are known as altitude grasslands and campos 1

rupestres (Vasconcelos 2011). The Pampa extends through Uruguay and Argentina and it is 2

assumed to have a relative floristic and structural continuum throughout its distribution (Bilenca 3

and Miñarro 2004; Ferreira and Boldrini 2011). However, this has not been evaluated quantitatively 4

so far within Brazilian territory. Although our sampling does not allow extrapolation of floristic 5

results for the two entire biomes, we did sample ca. 17% of the regional grassland flora, and also 6

found structural patterns consistent with previous descriptive studies (e.g., Boldrini 2009; Boldrini 7

et al. 2009). 8

Sites between biomes shared only 15% of their species, and the number of shared species in 9

pairwise comparisons was slightly lower in Atlantic Forest sites (Figure 2). Also, the ordination 10

analysis performed with presence/absence data resulted in a pattern of biome separation similar to 11

the one found with cover data (data not shown). These results indicate that differences between 12

biomes are also dependent on species composition. Distribution of species richness and relative 13

cover values per family also differed between biomes (Figure 1). Poaceae species represented a 14

slightly higher cover value in Atlantic Forest sites, although they were more numerous in Pampa 15

sites. Asteraceae represented higher cover in Pampa sites, whereas Cyperaceae species richness and 16

cover was more than two times higher in Atlantic Forest sites (this last could be related to more 17

humid climate in the Atlantic Forest). Also, there was a high family turnover rate between biomes 18

in less representative families. A regional literature-based floristic review found similar patterns, 19

with the Pampa biome showing more exclusive species (Boldrini 2009). These differences in 20

species composition (Table 3 and Figure 1) reflect differences in vegetation structure (Table 2) and 21

community parameters (Table 1 and Figures 3 and 4). Differences in climatic variables such as 22

mean annual rainfall (Nimer 1990), altitude and geological and soil features (Streck et al. 2008) are 23

probably important variables that influence these differences. Also, areas from both biomes are 24

under the influence of different floristic contingencies (Boldrini and Longhi-Wagner 2011; Cabrera 25

and Willink 1980), which is also reflected in different species composition and vegetation structure. 26

Page 34: abordagem quali-quantitativa e funcional de vegetação campestre ...

35

Although pairwise comparisons of plant composition and cover between sites of the same 1

biome showed no differences, variation of some parameters was different between biomes. Most 2

structural parameters (Table 2) were more variable within the Pampa biome, as was Shannon 3

diversity (Table 1). Diversity profiles showed that diversity in Atlantic Forest sites was overall 4

slightly lower than in Pampa sites (Figure 3). Also, diversity among Pampa sites did not differ with 5

increasing alpha values, whereas it did among Atlantic Forest sites. Changes in alpha values 6

represent changes in sensitivity to abundant and rare species. Higher alpha values correspond to 7

diversity indexes that give more weight to abundant species (Tóthmérész 1995). Therefore, 8

differences in diversity between Atlantic Forest sites are related to the high dominance seen in these 9

sites (Table 3), and are also reflected in the higher aggregation of paddocks in comparison with 10

Pampa sites revealed in the ordination analysis (Figure 4). 11

The most striking structural differences in sites between biomes were mean vegetation 12

height, cover and bare soil (Table 2). Some of these structural variables turned out to be good 13

predictors of grassland plant community parameters. We found strong positive correlations between 14

bare soil and species richness (r=0.75), Shannon diversity (r=0.93) and evenness (r=0.77) and 15

between rock outcrops and evenness (r=0.61). Also, mean vegetation cover and evenness were 16

negatively correlated (r=-0.63). Fire has been used to remove litter and standing dead biomass after 17

winter for years in Atlantic Forest grasslands (Jacques 2003). This historical practice, combined 18

with low stocking rates (Maraschin 2001; Nabinger et al. 2009) probably lead to the present 19

structure: high dominance of C4 grasses that form dense tussocks, such as Andropogon lateralis, 20

which accordingly grouped close to Atlantic Forest sites in the ordination analysis (Figure 4). Under 21

light grazing pressures, such tussock species are allowed to grow both in height and tussock 22

diameter, which may result in protection of growing buds, consequent resistance to future fire 23

events and contributes to maintain dominance (Jacques 2003; Overbeck et al. 2005). 24

Evidence found in grasslands from other parts of the world (at least under relatively humid 25

climate conditions) indicates that when disturbance (i.e. grazing and/or fire) is reduced or removed, 26

Page 35: abordagem quali-quantitativa e funcional de vegetação campestre ...

36

plant species richness and/or diversity also reduce (Altesor et al. 2005; Fynn et al. 2004; McIntyre 1

et al. 2003; Olff and Ritchie 1998; Overbeck et al. 2005; Rusch and Oesterheld 1997; Sebastià et al. 2

2008; Tremont 1994). Our results showed that sites with lower grazing pressures showed lower 3

values of richness and diversity (Table 1), and overall less exclusive species (Figure 2). Coupled 4

with the high correlation values we found between grazing pressure and richness, diversity and 5

vegetation height, our results suggest a close relationship between management and grassland 6

structure and composition. Management practices and the resulting high dominance are directly 7

linked to increased mean vegetation cover and height in Atlantic Forest sites (Table 2), which in 8

turn correlates to the lower values of richness and diversity. 9

High percentages of bare soil and rock outcrops are usually related to grasslands growing 10

over shallow soils (e.g., Lezama et al. 2011; Pinto et al. 2013). Plant communities that thrive on 11

such resource-limiting micro-environments are prone to show more evenly distributed abundances, 12

less dominance and consequently higher values of evenness and diversity (e.g., Pinto et al. 2013; 13

Setubal and Boldrini 2012). This explains the positive correlation between bare soil and rock 14

outcrops and richness, diversity and evenness. Although mean values for bare soil were different 15

between biomes and values for rock outcrop were not, they showed large variation between sites 16

(Table 2), and even higher between sampling units. Nonetheless, both variables were strongly 17

correlated to vegetation parameters, suggesting that soil features and vegetation structure are closely 18

related. This could be a reflection of soil effects on vegetation patterns at two levels. The first 19

depicts differences between biomes, mostly related to differences in soil acidity. The second shows 20

natural vegetation heterogeneity related to local topography, which was found to be a strong driver 21

structuring grassland communities at the landscape level (Sebastiá 2004). Relating abiotic and 22

biotic factors is not a new topic in ecology (Austin et al. 1990; Gibson et al. 1993; Grime 1979; 23

Tilman 1984). However, it is still used as a tool to explain grassland community organization (e.g., 24

Cantero et al. 2003; Fynn and O'Connor 2005), although the relative contribution of biotic, abiotic 25

and spatial factors in community assembly is still an open question in community ecology. 26

Page 36: abordagem quali-quantitativa e funcional de vegetação campestre ...

37

Diversity of soil types is higher in the Pampa biome when compared with the Atlantic Forest in 1

southern Brazil (Streck et al. 2008). Although our sampling does not allow for broader 2

extrapolations on soil-vegetation relationships, we did find an evident pattern linking abiotic and 3

biotic variables within our data (Figure 5). Also, correlation between axes containing soil and 4

vegetation information was significant and high (r=0.86), reinforcing the idea that vegetation 5

patterns are related to soil features. 6

The Rio de La Plata Grasslands, in which the Brazilian Pampa biome is inserted, are 7

characterized by a less seasonal environment in comparison with northern Hemisphere grasslands 8

(Paruelo et al. 1995). Most studies from these subtropical and temperate grasslands indicate the 9

predominance of prostrate growth forms (Rodríguez et al. 2003; Díaz et al. 1992 and our results for 10

grasslands in the Pampa biome). In Atlantic Forest grasslands, however, the dominance of tussock 11

species is similar to what is found in temperate grasslands such as North American prairies (Olff 12

and Ritchie 1998). Grazed grasslands under moderate to high grazing pressures are dominated by 13

prostrate growth forms, tend to accumulate less standing biomass, and are less prone to regular 14

burning in comparison with grasslands dominated by erect tussock species (Altesor et al. 2005; 15

Guerschman and Paruelo 2005). These differences in structure and management pose different 16

challenges for conservation. In our personal observations during fieldwork over the past years, we 17

have seen systematic substitution of natural grasslands, previously used as forage sources, by 18

croplands and exotic tree plantations. Transformation of natural ecosystems to croplands represents 19

one of the greatest threats to global biodiversity (Sala et al. 2000), and present conservation status 20

of grasslands in the region is probably much worse than presently estimated, since the study by 21

Cordeiro and Hasenack (2009) is based on remote sensing images from ten years ago. 22

According to current Brazilian environmental law, the use of fire or domestic herbivores is 23

not allowed inside conservation units under the most restrictive categories, which supposedly are 24

the ones that should provide the highest protection for biodiversity (MMA 2000; Olson and 25

Dinerstein 1998). Herbivores are used as management tools in conservation of grasslands in North 26

Page 37: abordagem quali-quantitativa e funcional de vegetação campestre ...

38

America (Meagher 1973) and Europe (Hampicke and Plachter 2010), among other places. 1

Furthermore, evidence of positive influence of grazing on ecosystems processes was found in 2

experiments carried out inside conservation units (e.g., Frank et al. 2000; Frank and McNaughton 3

1993). To create or maintain conservation units encompassing large natural grassland landscapes 4

that would remain unmanaged would be to repeat studies carried out worldwide that showed that 5

management exclusion leads to species loss and decreasing diversity. In fact, the vast majority of 6

preserved grassland ecosystems in southern Brazil are natural pasturelands used for extensive cattle 7

breeding (Cordeiro and Hasenack 2009). Grasslands we sampled in the Atlantic Forest were 8

inserted in conservation units, but as the areas were incorporated into these quite recently, grazing 9

has not yet been excluded – but current management plans aim to do so. Our results indicated that 10

grassland areas under lower grazing pressures harbor less species richness and diversity in 11

comparison with more heavily grazed sites. It is likely that the complete exclusion of management 12

would promote further biodiversity loss in these areas. If the prevailing Brazilian conservation 13

policy is maintained, grazing and fire will indeed be suppressed from these grasslands (and many 14

others throughout southern Brazil). This will ultimately lead, as discussed above, to declines in 15

plant richness and diversity, besides negative effects on richness and diversity in other trophic 16

levels as well as on ecosystem processes and services. 17

Our results indicated that grasslands in the Pampa and Atlantic Forest biomes differ from 18

one another considering plant species composition and vegetation structure, in consequence both of 19

environmental conditions (soil and climate) and current and past management. The implications of 20

this conclusion for conservation are immediate: conservation efforts must be equally focused on 21

grasslands from both biomes in order to target: (1) a set of species representative of the different 22

grassland types; (2) different vegetation structures and (3) potentially different ecological processes 23

and ecosystem services. Also, it is imperative that management is taken into account when planning 24

future conservation efforts focusing on natural subtropical grasslands. Management is important not 25

only due to its effects on biodiversity, but also because it is related to local culture and legal issues. 26

Page 38: abordagem quali-quantitativa e funcional de vegetação campestre ...

39

The next step to build a framework for conservation of grasslands in southern Brazil would be: (i) 1

to bridge the gap between farmers and scientists, in order to provide the first the sustainable 2

management alternatives that allow for simultaneous conservation and monetary gain and (ii) to 3

propose standardized protocols for implementation and long-term maintenance of a disturbance 4

regime in conservation units encompassing grasslands. Ongoing research projects in which our 5

work was included aim to provide further contribution to build and thread these steps. 6

7

Acknowledgements 8

9

We thank the environmental authorities (Secretaria Estadual do Meio Ambiente – São 10

Francisco de Paula and Instituto Chico Mendes de Conservação da Biodiversidade – Parque 11

Nacional dos Aparados da Serra) and farmers that made this work possible, and all those that helped 12

us on fieldwork and taxonomic identification of biological material. The first two authors thank the 13

Coordenação de Aperfeiçoamento de Pessoal de Nível Superior (CAPES) for scholarships. This 14

study was part of a long-term ecological research project (LTER/PELD Campos Sulinos; CNPq 15

558282/2009-1). 16

Page 39: abordagem quali-quantitativa e funcional de vegetação campestre ...

40

Tables

Table 1 Mean species richness, diversity (Shannon’s H’) and respective standard deviations for

grassland plant communities in six sites distributed in two biomes. Different letters represent

significant differences (at paddock level) within columns and borders (P<0.05). P = Pampa sites,

AF = Atlantic Forest sites

Biome Site Species richness Diversity (H')

Mean SD Mean SD

P 27.17ª 6.24 2.20ª 0.35

AF 22.86b 4.81 1.73b 0.34

ACE 33.04ª 5.18 2.43ª 0.28

P ALE 23.19b 3.70 2.20b 0.34

LAV 25.30b 4.85 1.98c 0.29

APA 25.63ª 5.58 1.83ª 0.21

AF ARA 21.96b 3.72 1.62ª 0.36

TAI 21.00b 3.71 1.74ª 0.40

Table 2 Mean values of structural variables for grassland plant communities in six sites distributed

in two biomes from Southern Brazil. Different letters represent significant differences within

biomes (P<0.05). P = Pampa sites, AF = Atlantic Forest sites

Biome Site Vegetation

cover (%)

Bare soil

(%)

Rock outcrop

(%)

Litter

(%)

Vegetation height

(cm)

P 80.43a 11.85a 1.42a 27.65a 8.70a

AF 91.91b 2.38b 1.05a 14.10a 27.64b

P

ACE 89.07a 13.15a - 25.56a 9.53a

ALE 75.74b 13.98a,b 4.26 32.78a 5.50b

LAV 76.48b 8.43b - 24.63a 11.06a

AF

APA 92.41a 2.50a 0.09a 19.89ª 24.73a,b

ARA 92.41a 3.15a,b 0.09a 20.09ª 24.07a

TAI 90.93a 1.48b 2.96b 2.31b 34.13b

Page 40: abordagem quali-quantitativa e funcional de vegetação campestre ...

41

Table 3 Relative cover (%) of the five most representative species per site of grassland

communities in six sites distributed in two biomes from Southern Brazil

Relative cover per site (%)

Pampa Atlantic Forest

Family Species ACE ALE LAV APA ARA TAI

Poaceae Andropogon lateralis - 6.98 0.04 40.11 48.15 33.81

Poaceae Axonopus affinis 3.59 0.54 6.39 1.52 4.87 0.94

Asteraceae Baccharis coridifolia - 4.85 4.56 - - -

Asteraceae Baccharis crispa 1.04 - 9.54 0.43 1.61 2.48

Cyperaceae Bulbostylis sp. - - - 5.87 - -

Poaceae Paspalum maculosum - - - 5.22 5.61 0.48

Poaceae Paspalum notatum 18.02 7.80 32.04 - - 3.96

Poaceae Paspalum plicatulum 7.93 - 0.39 0.87 0.02 1.13

Poaceae Piptochaetium montevidense 11.11 7.07 4.15 2.11 0.17 1.13

Cyperaceae Rhynchospora megapotamica 8.91 0.02 - - - -

Poaceae Saccharum angustifolium - - 1.50 - 3.52 -

Poaceae Schizachyrium tenerum - 0.02 0.02 4.98 1.06 13.13

Total 50.59 27.28 58.61 61.11 65.00 57.07

Page 41: abordagem quali-quantitativa e funcional de vegetação campestre ...

42

Figures

Fig. 1 Species richness, average cover per family and species per family ratio (number of species

divided by number of families multiplied by 100) in six grassland communities distributed in the

Pampa and Atlantic Forest biomes in Southern Brazil

Fig. 2 Venn diagrams showing shared species between grasslands from two biomes and among the

three sampling sites within each biome

Page 42: abordagem quali-quantitativa e funcional de vegetação campestre ...

43

Fig. 3 Diversity profiles of grassland communities from six sampling sites distributed in two

biomes in Southern Brazil. Alpha values represent Rényi entropy values

Page 43: abordagem quali-quantitativa e funcional de vegetação campestre ...

44

Fig. 4 Ordination diagram (Principal Coordinate Analysis) of six grassland communities (three

paddocks each) described by 382 variables (plant species). Plotted variables were the most

correlated with the first two ordination axes. Percentage of variation captured in both axes is shown

in parenthesis. Legend for variables: anla = Andropogon lateralis; axaf = Axonopus affinis; chsu =

Chascolytrum subaristatum; Comm = Commelina sp.; dise = Dichondra sericea; euas =

Eupatorium ascendens, hyde = Hypoxis decumbens; hyex = Hydrocotyle exigua; krfl =

Krapovickasia flavescens; kyva = Kyllinga vaginata; padi = Paspalum dilatatum; pano = Paspalum

notatum; pftu = Pfaffia tuberosa; pimo = Piptochaetium montevidense; plpe = Plantago penantha;

pter = Pteridophyta

Page 44: abordagem quali-quantitativa e funcional de vegetação campestre ...

45

Fig. 5 Linear regression model showing the relationship between ordination axes containing

information of vegetation (162 sampling units described by 382 plant species) and soil (162

sampling units described by 17 soil variables) from six grassland communities distributed in two

biomes

Page 45: abordagem quali-quantitativa e funcional de vegetação campestre ...

46

Electronic supplementary material

ESM 1 Ordination diagram (Principal Coordinate Analysis) of six grassland communities (27

sampling units each) described by soil variables. Legend for variables: SMP = SMP pH method;

Bases = percent base saturation; pH = pH (H2O); Ca/Mg = Ca:Mg ratio; Mg/K = Mg:K ratio; K =

potassium; MgT = exchangeable magnesium; CaT = exchangeable calcium; Phos = phosphorus;

Argi = clay content; MO = organic matter; CTC = cation exchange capacity; AlT = exchangeable

aluminum; Al = percent aluminum saturation; AlH = hydrogen-ion and aluminum concentrations

Page 46: abordagem quali-quantitativa e funcional de vegetação campestre ...

47

References

Altesor A, Oesterheld M, Leoni E, Lezama F, Rodriguez C (2005) Effect of grazing on

community structure and productivity of a Uruguayan grassland. Plant Ecology 179 (1):83-91

Aronson J, Milton SJ, Blignaut JN, Clewell AF (2006) Nature conservation as if people

mattered. Journal for Nature Conservation 14 (3):260-263

Austin M, Grace J, Tilman D (1990) Community theory and competition in vegetation. In: Grace

JB, Tilman D (eds) Perspectives on Plant Competition. Academic Press Inc., London, pp 215-238

Behling H, Pillar VD, Bauermann SG (2005) Late Quaternary grassland (Campos), gallery

forest, fire and climate dynamics, studied by pollen, charcoal and multivariate analysis of the São

Francisco de Assis core in western Rio Grande do Sul (southern Brazil). Review of Palaeobotany

and Palynology 133 (3):235-248

Behling H, Pillar VDP, Orlóci L, Bauermann SG (2004) Late Quaternary Araucaria forest,

grassland (Campos), fire and climate dynamics, studied by high-resolution pollen, charcoal and

multivariate analysis of the Cambará do Sul core in southern Brazil. Palaeogeography,

Palaeoclimatology, Palaeoecology 203 (3):277-297

Bilenca D, Miñarro F (2004) Identificación de áreas valiosas de pastizal en las pampas y campos

de Argentina, Uruguay y sur de Brasil. Fundación Vida Silvestre Argentina, Buenos Aires

Boldrini II (1997) Campos do Rio Grande do Sul: caracterização fisionômica e problemática

ocupacional. Boletim do Instituto de Biociências 56:1-39

Boldrini II (2009) A flora dos campos do Rio Grande do Sul. Campos sulinos: conservação e uso

sustentável da biodiversidade Brasília: Ministério do Meio Ambiente:63-77

Boldrini II, Eggers L, Mentz LA, Miotto STF, Matzenbacher NI, Longhi-Wagner HM, Trevisan

R, Schneider AA, Setubal RB (2009) Flora. In: Boldrini II (ed) Biodiversidade dos campos do

Planalto das Araucárias. MMA, Brasília, pp 39-94

Page 47: abordagem quali-quantitativa e funcional de vegetação campestre ...

48

Boldrini II, Longhi-Wagner HM (2011) Poaceae no Rio Grande do Sul: diversidade, importância

na fisionomia e conservação. Ciência e Ambiente 42 (71-92)

Cabrera AL, Willink A (1980) Biogeografía de América latina. Serie de Biología Monografías

13

Cantero J, Liira J, Cisneros J, Gonzalez J, Nuñez C, Petryna L, Cholaky C, Zobel M (2003)

Species richness, alien species and plant traits in Central Argentine mountain grasslands. Journal of

Vegetation Science 14 (1):129-136

Cordeiro JLP, Hasenack H (2009) Cobertura vegetal atual do Rio Grande do Sul. In: Pillar VDP,

Müller SC, Castilhos ZMC, Jacques A (eds) Campos Sulinos – conservação e uso sustentável da

biodiversidade. MMA, Brasília, pp 285 - 299

Díaz S, Acosta A, Cabido M (1992) Morphological analysis of herbaceous communities under

different grazing regimes. Journal of Vegetation Science 3 (5):689-696

Ellenberg D, Mueller-Dombois D (1974) Aims and methods of vegetation ecology. Wiley New

York, NY,

Ferreira PMA, Boldrini II (2011) Potential Reflection of Distinct Ecological Units in Plant

Endemism Categories. Conservation Biology 25 (4):672-679

Frank DA, Groffman PM, Evans RD, Tracy BF (2000) Ungulate stimulation of nitrogen cycling

and retention in Yellowstone Park grasslands. Oecologia 123 (1):116-121

Frank DA, McNaughton SJ (1993) Evidence for the promotion of aboveground grassland

production by native large herbivores in Yellowstone National Park. Oecologia 96 (2):157-161

Fuhlendorf SD, Harrell WC, Engle DM, Hamilton RG, Davis CA, Leslie Jr DM (2006) Should

heterogeneity be the basis for conservation? Grassland bird response to fire and grazing. Ecological

Applications 16 (5):1706-1716

Fynn RW, Morris CD, Edwards TJ (2004) Effect of burning and mowing on grass and forb

diversity in a long‐term grassland experiment. Applied Vegetation Science 7 (1):1-10

Page 48: abordagem quali-quantitativa e funcional de vegetação campestre ...

49

Fynn RW, O'Connor TG (2005) Determinants of community organization of a South African

mesic grassland. Journal of Vegetation Science 16 (1):93-102

Gibson DJ, Seastedt T, Briggs JM (1993) Management practices in tallgrass prairie: large-and

small-scale experimental effects on species composition. Journal of Applied Ecology:247-255

Grime J (1979) Plant Strategies and Vegetation Processes. New York: John Wiley,

Guerschman JP, Paruelo JM (2005) Agricultural impacts on ecosystem functioning in temperate

areas of North and South America. Global and Planetary Change 47 (2):170-180

Hampicke U, Plachter H (2010) Livestock grazing and nature conservation objectives in Europe.

In: Large-scale livestock grazing. Springer, pp 3-25

Heywood VH, Iriondo JM (2003) Plant conservation: old problems, new perspectives.

Biological Conservation 113 (3):321-335

Hill MO (1973) Diversity and evenness: a unifying notation and its consequences. Ecology 54

(2):427-432

Hodgson J (1990) Grazing management. Science into practice. Longman Group UK Ltd.,

Hoekstra JM, Boucher TM, Ricketts TH, Roberts C (2005) Confronting a biome crisis: global

disparities of habitat loss and protection. Ecology Letters 8 (1):23-29

Houston D (1982) The northern Yellowstone elk: ecology and management.

IBGE (2004) Mapa da vegetação do Brasil e Mapa de Biomas do Brasil. Instituto Brasileiro de

Geografia e Estatística–IBGE

Jacques AVA (2003) A queima das pastagens naturais–efeitos sobre o solo ea vegetação.

Ciência Rural 33 (1):177-181

Larigauderie A, Mooney HA (2010) The Intergovernmental science-policy Platform on

Biodiversity and Ecosystem Services: moving a step closer to an IPCC-like mechanism for

biodiversity. Current Opinion in Environmental Sustainability 2 (1):9-14

Lezama F, Altesor A, León RJ, Paruelo JM (2006) Heterogeneidad de la vegetación en

pastizales naturales de la región basáltica de Uruguay. Ecología Austral 16 (2):167-182

Page 49: abordagem quali-quantitativa e funcional de vegetação campestre ...

50

Lezama F, Altesor A, Pereira M, Paruelo J, Altesor A, Ayala W, Paruelo J (2011) Descripción

de la heterogeneidad florística de las principales regiones geomorfológicas de Uruguay. Bases

ecológicas y tecnológicas para el manejo de pastizales Ed Altesor A, Ayala W y Paruelo JM INIA,

Serie FPTA (26):15-32

Londo G (1976) The decimal scale for relevés of permanent quadrats. Vegetatio 33 (1):61-64

Magurran AE (1988) Ecological diversity and its measurement, vol 168. Princeton University

Press, Princeton

Maraschin GE Production potential of South American grasslands. In: XIX International

Grassland Congress, São Paulo, Brazil., 2001.

Mares MA (1986) Conservation in South America: problems, consequences, and solutions.

Science 233 (4765):734-739

McIntyre S, Heard K, Martin TG (2003) The relative importance of cattle grazing in subtropical

grasslands: does it reduce or enhance plant biodiversity? Journal of Applied Ecology 40 (3):445-

457

Meagher MM (1973) The bison of Yellowstone National Park. US Government Printing Office,

Milchunas D, Sala O, Lauenroth WK (1988) A generalized model of the effects of grazing by

large herbivores on grassland community structure. American Naturalist:87-106

Millennium Ecosystem Assessment (2005) Ecosystems and human well-being, vol 5. Island

Press Washington, DC,

MMA (2000) Sistema Nacional de Unidades de Conservação. vol Law no. 9.985. Ibama/MMA,

Brasília

Nabinger C, Ferreira E, Freitas A, Carvalho P, Sant’Anna D (2009) Produção animal com base

no campo nativo: aplicações de resultados de pesquisa. Campos sulinos: conservação e uso

sustentável da biodiversidade Brasília: Ministério do Meio Ambiente:175-198

Page 50: abordagem quali-quantitativa e funcional de vegetação campestre ...

51

Nabinger C, Moraes Ad, Maraschin G (2000) Campos in southern Brazil. Grassland

ecophysiology and grazing ecology:355-376

Nimer E (1990) Clima. In: IBGE (ed) Geografia do Brasil: Região Sul, vol 2. SERGRAF/IBGE,

Rio de Janeiro, pp 151-187

Olff H, Ritchie ME (1998) Effects of herbivores on grassland plant diversity. Trends in Ecology

& Evolution 13 (7):261-265

Olson DM, Dinerstein E (1998) The Global 200: a representation approach to conserving the

Earth’s most biologically valuable ecoregions. Conservation Biology 12 (3):502-515

Orloci L (1967) An agglomerative method for classification of plant communities. The Journal

of Ecology:193-206

Overbeck GE, Müller SC, Fidelis A, Pfadenhauer J, Pillar VD, Blanco CC, Boldrini II, Both R,

Forneck ED (2007) Brazil's neglected biome: The South Brazilian Campos. Perspectives in Plant

Ecology, Evolution and Systematics 9 (2):101-116

Overbeck GE, Müller SC, Pillar VD, Pfadenhauer J (2005) Fine‐scale post‐fire dynamics in

southern Brazilian subtropical grassland. Journal of Vegetation Science 16 (6):655-664

Paruelo J, Lauenroth W, Epstein H, Burke I, Aguiar M, Sala O (1995) Regional climatic

similarities in the temperate zones of North and South America. Journal of Biogeography:915-925

Pillar V (1997) Multivariate exploratory analysis and randomization testing with MULTIV.

Coenoses 12:145-148

Pillar V (1998) Sampling sufficiency in ecological surveys. Abstracta Botanica 22:37-48

Pillar V, Vélez E (2010) Extinção dos Campos Sulinos em unidades de conservação: um

fenômeno natural ou um problema ético. Natureza & Conservação 8:84-88

Pillar VD (1999a) The bootstrapped ordination re‐examined. Journal of Vegetation Science 10

(6):895-902

Pillar VD (1999b) How sharp are classifications? Ecology 80 (8):2508-2516

Page 51: abordagem quali-quantitativa e funcional de vegetação campestre ...

52

Pillar VDP, Orlóci L (1996) On randomization testing in vegetation science: multifactor

comparisons of relevé groups. Journal of Vegetation Science 7 (4):585-592

Pillar VDP, Quadros FLF (1997) Grassland-forest boundaries in southern Brazil. Coenoses 12

(23):119-126

Pinto MF, Nabinger C, Boldrini II, Ferreira PMDA, Setubal RB, Trevisan R, Fedrigo JK,

Carassai IJ (2013) Floristic and vegetation structure of a grassland plant community on shallow

basalt in southern Brazil. Acta Botanica Brasilica 27 (1):162-179

Quadros FLF, Pillar VDP (2001) Dinâmica vegetacional em pastagem natural submetida a

tratamentos de queima e pastejo. Ciência rural Santa Maria Vol 31, n 5 (set/out 2001), p 863-868

R Development Core Team (2012) R: A Language and Environment for Statistical Computing.

R Foundation for Statistical Computing. http://www.R-project.org/. 2013

Ramirez-Villegas J, Jarvis A, Touval J (2012) Analysis of threats to South American flora and

its implications for conservation. Journal for Nature Conservation 20 (6):337-348

Rodríguez C, Leoni E, Lezama F, Altesor A (2003) Temporal trends in species composition and

plant traits in natural grasslands of Uruguay. Journal of Vegetation Science 14 (3):433-440

Rusch GM, Oesterheld M (1997) Relationship between productivity, and species and functional

group diversity in grazed and non-grazed Pampas grassland. Oikos:519-526

Sachs JD, Baillie JE, Sutherland WJ, Armsworth PR, Ash N, Beddington J, Blackburn TM,

Collen B, Gardiner B, Gaston KJ (2009) Biodiversity conservation and the millennium development

goals. Science 325 (5947):1502-1503

Sala OE, Chapin FS, Armesto JJ, Berlow E, Bloomfield J, Dirzo R, Huber-Sanwald E,

Huenneke LF, Jackson RB, Kinzig A (2000) Global biodiversity scenarios for the year 2100.

Science 287 (5459):1770-1774

Sebastiá M-T (2004) Role of topography and soils in grassland structuring at the landscape and

community scales. Basic and Applied Ecology 5 (4):331-346

Page 52: abordagem quali-quantitativa e funcional de vegetação campestre ...

53

Sebastià MT, Bello F, Puig L, Taull M (2008) Grazing as a factor structuring grasslands in the

Pyrenees. Applied Vegetation Science 11 (2):215-222

Setubal RB, Boldrini II (2012) Phytosociology and natural subtropical grassland communities on

a granitic hill in southern Brazil. Rodriguésia 63 (3):513-524

Silva F (1999) Manual de análises químicas de solos, plantas e fertilizantes. Brasília:

EMBRAPA Comunicação para transferência de tecnologia

Soriano A, León R, Sala O, Lavado R, Deregibus V, Cauhépé M, Scaglia O, Velázquez C,

Lemcoff J (1992) Río de la Plata Grasslands. In ‘Ecosystems of the world 8A. Natural grasslands.

Introduction and Western Hemisphere’.(Ed. RT Coupland) pp. 367–407. Elsevier: Amsterdam,

Streck EV, Kämpf N, Dalmolin RS, Kamt E, Nascimento Pd, Schneider P, Giasson E, Pinto L

(2008) Solos do Rio Grande do Sul. UFRGS, Departamento de Solos, Faculdade de Agronomia,

Tilman GD (1984) Plant dominance along an experimental nutrient gradient. Ecology:1445-

1453

Tóthmérész B (1995) Comparison of different methods for diversity ordering. Journal of

Vegetation Science 6 (2):283-290

Tremont R (1994) Life-history attributes of plants in grazed and ungrazed grasslands on the

Northern Tablelands of New South Wales. Australian Journal of Botany 42 (5):511-530

Vasconcelos MF (2011) O que são campos rupestres e campos de altitude nos topos de montanha

do leste do Brasil? Revista Brasileira de Botânica 34 (2):241-246

Page 53: abordagem quali-quantitativa e funcional de vegetação campestre ...

54

Page 54: abordagem quali-quantitativa e funcional de vegetação campestre ...

55

Plant life forms revisited: are classic systems really applicable in all ecosystems?

Pedro M.A. Ferreira, Gerhard E. Overbeck, Ilsi I. Boldrini

P.M.A. Ferreira (corresponding author)

Universidade Federal do Rio Grande do Sul, Programa de Pós Graduação em Botânica, Av. Bento

Gonçalves 9500 Bloco IV, P. 43432, CEP 91501-970, Porto Alegre, RS, Brazil.

e-mail: [email protected]

phone: +55 51 3308 7555

Page 55: abordagem quali-quantitativa e funcional de vegetação campestre ...

56

Abstract 1

2

Plant life forms are coarse classifications of taxonomic entities used to describe biological 3

patterns and processes. Such classifications are present in science since its early days, and are still 4

used as descriptive tools in areas such as functional ecology and biogeography. Some classifications 5

became widely accepted and applied worldwide. In this paper we present a review of plant life form 6

classifications. We discuss the relevance and accuracy of using such all-encompassing 7

classifications in any given ecosystem to answer varying ecological questions. We propose a multi-8

trait hierarchical classification of plant life forms for subtropical grasslands, using a case study from 9

southern Brazil to compare its descriptive power with the widely used Raunkiaer and Ellenberg & 10

Mueller-Dombois classifications. To perform such comparisons we used Mantel tests and 11

Procrustes analyses in a multivariate matrix-based approach. Subtropical grassland species were 12

mostly grouped into large heterogeneous categories in classic one-trait-based life form systems. 13

These systems showed poor descriptive power of differences between areas with known differences 14

in management, environmental factors and vegetation structure. Our classification was a better 15

descriptor of species-based patterns in comparison with the existing ones. Our results indicated that 16

the descriptive power of a life form classification lies in the criteria used to group species rather 17

than in the number of categories alone. We pointed out that using a life form classification 18

consistently within an ecological unit may reflect in future benefits, such as facilitating and 19

improving the accuracy of meta-analyses and allowing the development of unified databases. This 20

might be particularly important considering transnational biomes such as the subtropical and 21

temperate South American grasslands. Although our classification was conceived in and for 22

subtropical grasslands, we suggest that researchers using life forms as descriptive traits in other 23

ecosystems consider the relatedness between the principles underlying the classification and the 24

ecological question being addressed, as well as the ecosystem in question. 25

26

Keywords: Raunkiaer system, campos grasslands, growth forms, traits 27

28

Page 56: abordagem quali-quantitativa e funcional de vegetação campestre ...

57

Introduction 1

2

Scientists have relied on coarse classifications of taxonomic entities to describe biological 3

patterns and processes since the dawn of ecology as a science. Earth’s biodiversity is simply too 4

complex to be described using the specific level (Díaz and Cabido 1997; Walker 1992), and many 5

ecological processes and patterns are better described by non-taxonomic units, although 6

phylogenetic similarities between these units are also important (Westoby 2006). In plant ecology, 7

classifications of life forms were extensively used from the descriptive works of the early 8

nineteenth century until today in functional approaches. Many classification systems were proposed 9

during all that time, although some became more accepted and widely used than others. Also, the 10

applicability of life form classifications conceived for universal use was criticized in some 11

ecosystems. In this paper we present a historical review of the use of plant life form classifications. 12

We question the use of widely accepted classifications indiscriminately in any ecosystem, 13

sometimes ignoring the relatedness between the principles underlying the classification and the 14

ecological question being addressed. We also propose a classification of plant life forms for 15

subtropical grasslands, and use a case study from southern Brazilian grasslands to compare it with 16

two widely used classifications. 17

18

Brief history of plant life form classifications 19

20

During the past century, the name of the Danish botanist Christen Raunkiaer became nearly 21

a metonym for plant life forms. However, grouping plant taxa into coarse and often taxonomy-22

independent classifications has been done for many years before his widely used classification 23

(Raunkiaer 1934). The first known classification of life forms dates back to ca. 300 BC, to 24

Theophrastus’ ‘Enquiry into plants’, in which the Greek philosopher defined four plant ‘classes’ 25

that scientists have been using ever since: tree, shrub, under-shrub and herb (Hort 1916). Although 26

Page 57: abordagem quali-quantitativa e funcional de vegetação campestre ...

58

science has moved from that exclusively descriptive approach, we still strive to find meaningful 1

working units in plant ecology, and often go back to Theophrastus’ four classes or minute variations 2

of them. 3

The scientific literature of the nineteenth century teems with examples of plant 4

classification systems, most of which comprising life forms based on physiognomic aspects. This 5

focus on physiognomy arose from the need to define primary units for the description of plant 6

communities (Du Rietz 1931). The system proposed by Humboldt (1806) was widely used by 7

European botanists during the first half of that century. The alternative system of De Candolle 8

(1818) was based on life span, and also included morphological features on its bases, thus being the 9

only system not exclusively based on physiognomy at the time. Humboldt’s system was 10

subsequently extended by Grisebach (1872), resulting in 54 ‘physiognomic types’. By that time, the 11

evolutionist paradigm (Darwin 1859) was already an influence, and plant classification systems 12

slowed shifted their focus from physiognomy to characters reflecting biological importance (e.g., 13

Kerner 1869). In fact, in his review of plant life forms, Du Reitz pointed out that ‘it was probably 14

only due to the personal influence of the strongly antievolutionist old Grisebach, that purely 15

physiognomic systems of vegetation-forms remained predominating more than two decades after 16

the publication of The Origin of Species’ (Du Rietz 1931, p. 4). 17

The period of the late nineteenth and early twentieth century was marked by the transition 18

between physiognomic and biological plant classification systems. Reiter (1885) was a pioneer by 19

proposing a revision of the classic physiognomic systems under an evolutionary view. However, 20

most of the systems conceived during this period were based on the Lamarckian concept of 21

epharmony, in which the environment would directly induce transformations in individuals (Vesque 22

1882). Near the turn of the century, Warming (1895) published his first attempt of classification of 23

plants into biological groups, mostly based on the work of De Candolle (1818), but including many 24

additional morphological features and for the first time using the term ‘life-form’. Warming 25

progressively refined his system in the following years, and synthetized his views in his book 26

Page 58: abordagem quali-quantitativa e funcional de vegetação campestre ...

59

‘Oecology of Plants’ (Warming and Vahl 1909). In that book, Warming introduced the English term 1

‘growth-form’, although the corresponding terms in Danish and German from his previous works 2

were ‘livsform’ and ‘Lebensform’ respectively. As Du Rietz (1931) argued later, he did so ‘without 3

giving any reason for not using the term life-form in English’. Until today, life form and growth 4

form are sometimes treated as synonyms. Oscar Drude, a severe critic of Humboldt’s and 5

Grisebach’s systems, also proposed a ‘biological system’ (Drude 1896), upgraded in further 6

publications and relatively influential at the time , being the foundation for some similar systems 7

that followed. In the same year, Areschoug (1896) used for the first time the term ‘geophyte’ for 8

plants with belowground renewal buds. 9

The first accounts of Raunkiaer’s life-form system are from publications in Danish and 10

French between 1904 and 1907 (Raunkiaer 1904; 1905; 1907). In the following decades, the system 11

underwent several changes (such as the increased number of species used in the calculation of the 12

‘normal spectra’) until the publication of ‘The life forms of plants and statistical plant geography’ 13

(Raunkiaer 1934). Although Raunkiaer’s system became widely accepted and repeatedly used 14

afterwards, not all his contemporary fellow scientists agreed with his views, and many kept on 15

updating their systems or developing new ones (e.g., Sylvén 1906). Warming was probably one of 16

his most prominent critics, but by far not the only one (see a review in Du Rietz 1931). One of the 17

most common criticisms (or contributions) to Raunkiaer’s system was directed towards the 18

establishment of consistent differences between Hemicryptophytes and Chamaephytes, as well as 19

the wide diversity of life forms that both classes actually encompassed (e.g., Skottsberg 1929; 20

Skottsberg 1913). 21

In 1920, the North American botanist Frederic Edward Clements published a new system, 22

restoring the term ‘vegetation-form’ in place of life form (Clements 1920). For Clements, the 23

dominant species played the central role in ecological processes, and thus their life forms were the 24

ones that mattered. Following Clements’ work, in 1921 the Swedish biologist Gustaf Einar Du Rietz 25

published his system, shifting back to an almost purely physiognomic-based classification (Du 26

Page 59: abordagem quali-quantitativa e funcional de vegetação campestre ...

60

Rietz 1921). Du Rietz was the first to establish a distinction between life and growth forms: the first 1

was determined by physiognomy alone, whereas the latter encompassed subdivisions based on 2

shoot architecture. Kerner’s system was probably the first one purely based on morphology, and 3

theoretically independent from taxonomy (Kerner 1929). In 1928 Josias Braun-Blanquet presented a 4

classification based on Raunkiaer’s, but including radically changed subdivisions of the original ten 5

main categories (Braun-Blanquet 1928). Ellenberg and Mueller Dombois (1967) presented a 6

refinement of Raunkiaer’s system using finer categories (later re-edited in Ellenberg and Mueller 7

Dombois 1974). It is important to note that, as Adamson (1939) pointed out, most of these 8

classifications of life forms are solely based on the aerial parts of the plants, completely ignoring 9

belowground differences. 10

After this period of copious production of plant classification systems, scientists moved 11

towards progressively more complex questions, often involving relationships between multiple 12

taxonomic levels and ecological processes. However, to answer many of these new questions, 13

classifications of plant taxa into simple non-taxonomic entities, such as life forms, were (and still 14

are) used. The most obvious example lies in functional ecology, which largely relies on recurrent 15

relationships between species traits such as life form and ecosystem function (e.g., Díaz and Cabido 16

1997; McIntyre et al. 1995). Although it is known that some ‘core traits’ are better descriptors of 17

ecological processes and environmental filters (Weiher et al. 1999; Westoby 1998), easy-to-18

measure traits such as life forms may also be useful, especially considering processes related to 19

vegetation structure. Since Raunkiaer’s system was simple, widely accepted and used for a good 20

many years, it was the classification of choice for many researchers around the world until today. 21

But would it be wise to keep using a classification system based solely on plant tolerance to a 22

generalized unfavorable season in all ecosystems, to answer a vast array of ecological questions? 23

24

25

26

Page 60: abordagem quali-quantitativa e funcional de vegetação campestre ...

61

Life forms based on single or multiple traits? 1

2

We showed above that the first distinction between life and growth forms may have arisen 3

from a language-related issue in Warming and Vahl (1909), as first noted by Du Rietz (1931). After 4

that, Du Rietz (1921) used the term growth form to describe morphology-based subdivisions of his 5

physiognomy-based life forms. Whittaker (1975) pointed out that life form (in Raunkiaer’s system) 6

considers only one characteristic (height of the perennating tissue in relation to ground level), 7

whereas growth form reflects a mixture of characteristics. Accordingly, recent protocols that aim to 8

standardize measurement of plant functional traits present growth and life forms as separate traits 9

(Cornelissen et al. 2003; Pérez-Harguindeguy et al. 2013). Be that as it may, we should keep in 10

mind that this distinction between life and growth forms considers ‘life form’ as in Raunkiaer’s 11

one-character system. However, as we have discussed above, there are many other life form 12

systems, most of which used multiple characters. Warming and Vahl (1909), for example, defined 13

life/growth forms as the sum of adaptive characters in a species expressing the relationship between 14

a plant and its environment. Although this definition may sound slightly outdated, it implies the 15

multi-character approach used in many modern functional studies, in which sets of traits are used to 16

group species into functional types to assess ecological processes in multiple organization levels. In 17

this view, life forms may be based on a set of traits, which may include plant features usually 18

related to ‘growth forms’, such as canopy structure. Therefore, from this point on we will use the 19

term life form consistently, and will no longer use the term ‘growth form’, to avoid further 20

confusion. 21

22

Plant life forms as predictors of ecological processes 23

24

‘Structure without function is a corpse; function without structure is a ghost.’ 25

(Vogel and Ewel 1972) 26

Page 61: abordagem quali-quantitativa e funcional de vegetação campestre ...

62

1

Under a functional perspective, plant life forms can be considered as functional groups 2

based on a single character (in Raunkiaer’s system; Solbrig 1993) or on multiple characters. Ewel 3

and Bigelow (1996) postulated that, at least in tropical ecosystems, it is the diversity of life forms, 4

and not species, that exerts major control over ecosystem functioning. For these statements to be 5

true, life forms must to some degree reflect the ecological processes in question. Therefore, by 6

using Raunkier’s life forms, we assume that the relative position of growing buds is a functionally 7

important character, and that his ten classes accurately represent the ecosystem being studied. 8

Although this has been shown to hold true in many studies around the world (regions with a more or 9

less clearly defined unfavorable season), not all researchers that used Raunkiaer’s life forms did so 10

without questioning. Already in the early 1900’s, Skottsberg (1913) pointed out that plants from the 11

Falklands classified as Hemicryptophytes actually behaved like Chamaephytes and, moreover, that 12

they were in fact evergreen plants. More recently, Pillar and Orlóci (1993) were also skeptical 13

regarding the use of assumedly universal classifications in any ecological context. 14

The applicability of Raunkiaer’s system in tropical and subtropical ecosystems has been 15

questioned mostly for two reasons: (i) it groups species in very few and apparently too broad, 16

uninformative categories (e.g., Ewel and Bigelow 1996) and (ii) an unfavorable season is hardly a 17

limiting factor in most tropical and subtropical realities, and Raunkiaer’s whole system is based on 18

plant resilience to this factor (Sarmiento and Monasterio 1983). In his studies on New Zealand 19

vegetation, Allan (1937) found out that Raunkiaer’s life form spectra did not actually reflect 20

climatic conditions. He concluded that the delimitations of Hemicryptophytes and Chamaephytes 21

were inadequate for austral floras, and that historical factors (i.e., the evolutionary history of the 22

studied ecosystem) should be taken into account when classifying plants into life forms. Adamson 23

(1939) pointed out that most classic plant life form classifications were based on floras from the 24

North Temperate Zone, with limited or no confirmatory evidence from Southern Europe, let alone 25

from the Tropics or Southern Hemisphere subtropical and temperate ecosystems. Considering 26

Page 62: abordagem quali-quantitativa e funcional de vegetação campestre ...

63

grassland ecosystems, Willems (1985) argued that alternatives to Raunkiaer’s system should be 1

used, since most species from grasslands (in the Netherlands) would be classified as 2

Hemicryptophytes, including obviously different life forms such as grasses and forbs in the same 3

class, thus obscuring shifts in community dynamics unraveled by his simpler classification. 4

Considering grassland ecosystems, there are many examples of case-specific life form 5

classifications that proved to be useful. Arnold (1955) presented a very simple classification based 6

on life span and height to evaluate rangeland condition. He found that range productivity was 7

dependent on the prevalent life forms according to his classification, which proved to be 8

informative also in the evaluation of ecological dominance and susceptibility to grazing. Willems 9

(1985) distinguished grassland plants in graminoids, forbs, rosettes and woody species, with some 10

sub-categories and ten final life forms. Using this classification to compare different management 11

regimes, he was able to establish relationships between biomass production and dominance, 12

vegetation structure and diversity. In African Savannas, Cramer et al. (2012) discussed the role of 13

underground competition in the coexistence of trees and grasses using the simplest life form 14

classification possible in their case: trees and grasses. 15

Further examples of case-specific classifications and criticism on classic classifications of 16

plant life forms are plentiful. However, our point here is that any classification of plants into a 17

system of life forms should be based on two principal aspects: (i) the evolutionary history of the 18

studied ecosystem and historical disturbance regimes it was submitted to, and (ii) the ecological 19

question the classification is being used to answer. Classic and all-encompassing systems such as 20

that of Raunkiaer may be useful to answer questions on a coarse scale, such as establishing broad 21

differences between floras of different regions. However, when searching for more specific 22

ecological processes that show little relatio to the theory underlying the classic life form system 23

proposed by Raunkiaer, they should be adapted, or independent classifications should be used 24

according to each case (as suggested by Allan 1937). In fact, examples of the use of case-specific 25

life form classifications from various ecosystems and involving various questions are abundant in 26

Page 63: abordagem quali-quantitativa e funcional de vegetação campestre ...

64

the recent literature (Aronson et al. 2007; Bhattarai and Vetaas 2003; Campanella and Bertiller 1

2008; Castanho et al. 2012; Collins and Calabrese 2012; Diaz et al. 2007; Gómez-Aparicio 2009; 2

Hadar et al. 1999; Huang et al. 2009; Ivanova 2012; Lezama et al. 2013; López et al. 2013; 3

Moustakas et al. 2013; Nelis 2012; Pekin et al. 2012; Skaer et al. 2012; Tsujino and Yumoto 2013). 4

5

Life forms in subtropical and temperate grasslands 6

7

In Southern South America, grasslands extend over ca. 750,000 km2 in Argentina, Uruguay 8

and Southern Brazil (Bilenca and Miñarro 2004; Soriano et al. 1992). These ecosystems evolved 9

under different levels of grazing (Milchunas et al. 1988) and fire (Behling et al. 2005; Behling et al. 10

2004). They have been used as a natural forage source for cattle breeding since the seventeenth 11

century (Pillar and Quadros 1997), and fire is still used as a management tool in some regions 12

(Jacques 2003). Therefore, a classification of life forms that is ecologically meaningful for these 13

ecosystems should take into account this historical disturbance regime. Life form categories should 14

be based on characters potentially related to plant adaptations to fire and herbivory. 15

Previous studies carried out in South American grasslands have used different life form 16

classifications. Some of them used adaptations based on Raunkiaer’s systems (e.g., Garcia et al. 17

2002; Overbeck and Pfadenhauer 2007). Other studies used case-specific classifications (e.g., 18

Altesor et al. 2006; Lezama et al. 2006; Overbeck et al. 2006; Overbeck et al. 2005). In grasslands 19

in the region, perennial species are predominant over annual species, and this ratio gradually 20

changes towards higher latitudes and temperate climate (Burkart 1975). Raunkiaer's system is 21

capable of detecting this change: the contribution of therophytes (frequency, cover and/or number 22

of species) would increase with latitude. 23

However, the problem of using Raunkiaer’s system (or any posterior adaptation based on it) 24

in these ecosystems is establishing the distinction between chamaephytes and hemicryptophytes. 25

Hemicryptophytes are defined as plants with ‘periodic shoot reduction to a remnant shoot system 26

Page 64: abordagem quali-quantitativa e funcional de vegetação campestre ...

65

that lies relatively flat on the ground surface’ (Ellenberg and Mueller-Dombois 1974; Ellenberg and 1

Mueller Dombois 1967; Raunkiaer 1934). This definition is still used and reproduced in many 2

works worldwide, including widely accepted protocols for measurement of plant traits (Cornelissen 3

et al. 2003; Pérez-Harguindeguy et al. 2013). Most plants from subtropical grasslands show no such 4

behavior, and would actually be more accurately classified as chamaephytes. This problem was 5

identified long ago for plants from the Falklands (Skottsberg 1913). In fact, by using Raunkiaer’s 6

original system, most of the non-therophytic grassland species from the subtropics would fall into 7

chamaephytes, thus creating a large and heterogeneous group. Such an all-encompassing category is 8

prone to show low descriptive power and to dim ecological patterns and processes. 9

In the following sections we propose a new classification of plant life forms for subtropical 10

grasslands, based on plant characteristics related to herbivory and fire. We test the robustness and 11

descriptive power of the new classification by comparing six grassland sites from Southern Brazil 12

with different management history and vegetation structure. Also, we compare this new 13

classification with the original system of Raunkiaer (1934) and the extended system proposed by 14

Ellenberg and Mueller Dombois (1967). 15

16

Material and Methods 17

18

We used data from a grassland vegetation survey carried out in 18 paddocks at six sampling 19

sites in Southern Brazil, three inserted in the Pampa Biome and three in the Atlantic Forest Biome. 20

Details on surveyed sites, sampling methods and results are described in Ferreira et al. (unpublished 21

[Capítulo 1]). Areas within each biome have been submitted to different management regimes in the 22

past decades (grazing and sporadic mowing in the Pampa and grazing and yearly fire in the Atlantic 23

Forest). Vegetation structure and community parameters differ between sites, mostly due to 24

management, soil features and environmental variables Ferreira et al. (unpublished [Capítulo 1]). It 25

is desirable that a classification of plant life forms also reflects these differences, since they are the 26

Page 65: abordagem quali-quantitativa e funcional de vegetação campestre ...

66

product of the ecosystem’s evolutionary history. Plant species from the survey with relative cover 1

value per site higher than five percent were classified in life forms according to Raunkiaer’s 2

original system (Raunkiaer 1934, henceforth mentioned as 'Raunkiaer's system'), the extended 3

system proposed by Ellenberg and Mueller Dombois (1967, henceforth mentioned as 'Ellenberg's 4

classification') and the new classification we propose in this paper. We used specific taxonomic 5

literature to classify each taxon. Species names were checked using The Taxonomic Name 6

Resolution Service (Boyle et al. 2013). 7

Since most plant species from subtropical grasslands do not present the ‘periodic shoot 8

reduction’ described for hemicryptophytes (i.e., they are evergreen plants), we considered that they 9

were actually chamaephytes in Raunkiaer’s system. We are aware that many previous studies that 10

used Raunkiaer’s system in similar ecosystems considered many species as hemicryptophytes. 11

Tussock-forming grasses, for example, are commonly considered hemicryptophytes because their 12

gems are slightly above the soil level, and often protected by the base of the tussock structure. 13

However, similar species from subtropical grasslands do not present a key feature needed to be 14

classified as hemicryptophytes: periodic shoot reduction, which implies they are evergreens. As 15

shown in the previous section, this distinction has already been pointed out by Skottsberg (1913) for 16

Argentinian temperate grasslands: most species actually behave like evergreen chamaephytes rather 17

than hemicryptophytes sensu Raunkiaer. Although there may be arguments supporting subtropical 18

and temperate species as hemicryptophytes, we will not delve further into this discussion in this 19

paper, and will consider species that do not show periodic shoot reduction as chamaephytes. In fact, 20

the distinction between categories does not affect our results and interpretations: we used 21

Raunkiaer’s classification for the sake of comparison with other systems (see below), and all 22

species that we classified as chamaephytes would remain under the same category regardless of the 23

name we use to present it. 24

To assess the predictive power of our classification and compare it with the classic ones, we 25

used a multivariate approach. First, we performed an ordination analysis (Principal Coordinate 26

Page 66: abordagem quali-quantitativa e funcional de vegetação campestre ...

67

Analysis) of a matrix containing the sampling areas described by species cover. We did the same 1

analysis using averaged cover values of life form categories as descriptors of sampling areas, 2

generating one ordination (sampling areas and life forms) per classification system. Based on the 3

original species-based matrix and the different life form classification systems, we generated four 4

matrices containing sampling units described by species and weighted by plant life forms. Since the 5

number of categories may affect the predictive power and resolution of classifications, we tested 6

two levels of our classification (five and 12 categories) to build two different life-form-weighted 7

matrices. We did the same using randomly generated life forms with the same number of categories 8

of all classifications. Methods to build trait-weighted matrices involve a series of matrix 9

multiplications described by Pillar et al. (2009) and Pillar and Duarte (2010). We used Mantel tests 10

and procrustes analyses (Peres-Neto and Jackson 2001) to evaluate the congruence between 11

variations in the generated life-form-weighted matrices and the original matrix of sampling areas 12

described by species cover. We performed all ordination analyses using chord distance as 13

dissimilarity measure between sampling units. We assessed differences of life forms between areas 14

from different biomes using randomization tests with 10,000 permutations (Pillar and Orlóci 1996). 15

All analyses were carried out using softwares Multiv (Pillar 1997) and on the R platform (R 16

Development Core Team 2012) with package ‘vegan’ (Oksanen et al. 2013). 17

18

Results 19

20

A classification of plant life forms for subtropical grasslands 21

22

We propose a hierarchical life form classification comprising four levels of progressively 23

refined categories (Figure 1, Table 1). The classic categories ‘therophytes’ and ‘geophytes’ are 24

maintained. However, Raunkiaer’s original chamaephytes are treated as ‘evergreens’, and then 25

progressively sorted into finer categories based on level of lignification, plant architecture, habit 26

Page 67: abordagem quali-quantitativa e funcional de vegetação campestre ...

68

and strategy of horizontal propagation (Table 1). In Electronic Supplementary Material 1 (ESM 1) 1

we provide a data matrix containing the life form categories for the 193 grassland plant species we 2

used in our analyses. 3

4

Comparison between classifications 5

6

Randomization tests showed significant differences (P<0.05) between areas from the 7

different biomes using our classification and Ellenberg’s, but not when using Raunkiaer system. 8

Therophytes, classified in the same way in all systems, showed significant higher cover values in 9

areas from the Pampa biome (P<0.01). Considering the classes from Ellenberg’s classification, 10

cover of herbaceous chamaephytes was significantly higher in Atlantic Forest areas (P<0.05). 11

Among the categories from our classification, we found significant differences between areas from 12

different biomes for solitary and connected tussocks, rhizomatous plants and stoloniferous plants 13

(P<0.01). The most prominent overall differences between biomes were seen in tussocks and 14

rhizomatous plants (Figure 2A). 15

Ordinations of matrices containing averaged cover values of life form categories resulted in 16

different patterns according to each classification (Figure 3). Therophytes were highly associated 17

with the second axis in all ordinations, segregating paddocks from site P1 from the others in varying 18

levels of clarity. Patterns of biome separation and within-site aggregation were more clearly 19

depicted by our classification (Figure 3A, B) in comparison with the other two systems. Using our 20

classification, sites between biomes were separated along axis 1. Atlantic Forest sites were more 21

associated with erect life forms, whereas Pampa sites P3 and P2 were more associated with 22

prostrate and ligneous life forms. This pattern was mostly lost when using the other two 23

classifications (Figure 3C, D). 24

Mantel tests and procrustes analyses revealed that matrices weighted by life form categories 25

from our classification (levels C and D) are more correlated to the species-cover matrix in 26

Page 68: abordagem quali-quantitativa e funcional de vegetação campestre ...

69

comparison with matrices weighted by Ellenberg’s and Raunkiaer’s classifications (Table 2 and 1

Figure 4). Matrices weighted by randomly generated life forms showed low and nonsignificant 2

correlation with species-described matrices. 3

4

Discussion 5

6

We have presented a new proposal of classification of plant life forms to be used in 7

subtropical and temperate grasslands. We have structured the classification in progressively refined 8

categories, so that it can be applicable at different levels of organization and scales (considering 9

both grain and extent). Also, we tested the descriptive power of our classification and compared it 10

with classic and broadly used classifications. 11

The second step in the system we proposed, in which we segregate ‘Geophytes’ from 12

‘Evergreens’ (Figure 1), deserves a brief discussion. The evergreens encompass the chamaephytes 13

from Raunkiaer’s classification (Table 1). Accordingly, in our view they are plants that do not have 14

their aerial parts reduced during part of the year due to climatic limitations - such limitations are not 15

common in subtropical grasslands. The geophytes, on the other hand, present this behavior, and are 16

naturally reduced to underground organs during part of the year, from which they resprout in the 17

next year. Although this adaptation may seem rather superfluous considering the present prevalent 18

climatic conditions in the subtropics and temperate zones, they may derive from past conditions. 19

During the late Pleistocene, grasslands were the dominant ecosystem in southern South America, 20

under a colder and dryer climate and submitted to grazing and burning events (Behling et al. 2005; 21

Behling et al. 2004; Bredenkamp et al. 2002; Milchunas et al. 1988). Therefore, this evolutionary 22

history may be partly responsible for the presence of geophytes in these ecosystems, and may lead 23

to a potential problem in segregating geophytes from evergreens: does the reduction in the aerial 24

structure occurs without external interference or is dependent on disturbance (grazing and/or fire)? 25

We considered that ‘true geophytes’ show the first behavior, i.e., reduction of above-ground shoots 26

Page 69: abordagem quali-quantitativa e funcional de vegetação campestre ...

70

is an inherent characteristic of their life cycle, whereas the second corresponds to evergreen plants 1

with underground storage organs that may allow post-disturbance resprouting (e.g., rhizomatous 2

evergreens). These evergreen resprouters may contribute to the resilience of these ecosystems to 3

historical disturbances (Fidelis et al. 2014; Fidelis et al. 2010). Although both life forms show 4

obviously similar adaptations to fire/grazing, geophytes are reduced to their underground organs 5

independently from disturbance events, possibly due to shared evolutionary history. Evergreens 6

with storage organs are more likely to respond to disturbance gradients (e.g., Overbeck and 7

Pfadenhauer 2007 and our results). Finally, we must consider that under extreme circumstances, 8

such as a unusally dry season or after a severe frost, both geophytes and evergreens are likely to 9

have their aboveground structures reduced. 10

Our classification (at levels C and D; Table 1 and Figure 1) was a better descriptor of the 11

original species-based patterns in comparison with other classifications (Table 2, Figures 2, 3 and 12

4). At the level C, our classification comprises five categories, the same number of Ellenberg’s 13

classification. Even so, the first showed better descriptive power than the latter (and also than 14

Raunkiaer’s system with four categories). Also, randomly generated life forms with the same 15

number of categories showed very limited descriptive power. This indicates that the descriptive 16

power lies in the criteria used to group species into life forms rather than in the number of 17

categories alone. Although this seems obvious, classic life form classifications have often been used 18

without considering their underlying classification criteria, which consequently may have led to 19

limitations regarding results and interpretations. For example, although differences in life forms 20

between grassland sites with different grazing levels have been repeatedly found using case-specific 21

classifications (e.g., Altesor et al. 2006; Diaz et al. 2007; Lezama et al. 2013), some authors have 22

found no such differences using Raunkiaer’s system (e.g., Vashistha et al. 2011). Our results 23

indicate that such differences could be captured by the life form categories we have proposed, 24

whereas coarse classifications such as that of Raunkiaer are not suited to answer specific ecological 25

questions such as the effect of different land management. Therefore, when the goal of a given 26

Page 70: abordagem quali-quantitativa e funcional de vegetação campestre ...

71

study is to sample general patterns, and not richness or diversity, these life form categories could 1

even be used in the sampling process in place of species. Considering that only within Brazilian 2

territory subtropical grasslands encompass more than 2,200 species (Boldrini 2009), this would 3

speed up sampling and allow larger areas to be covered, ultimately improving the still poorly known 4

general structural patterns of these ecosystems. However, sampling at such a coarse level and not 5

considering species’ identities would imply severe limitations, especially considering analytical 6

possibilities, including the possibility to refine or adapt the life form classification at a later point in 7

time. 8

In subtropical grasslands, Raunkiaer’s chamaephytes actually encompass very different 9

types of plants (Table 1), and using this system thus resulted in almost complete lack of differences 10

between biomes (Figure 2B), less informative ordination patterns (Figure 3C) and the lowest 11

congruence with species-based patterns (Figure 4). Previous works had already pointed out the 12

‘chamaephyte problem’, arguing that this category may actually encompass several life forms, 13

depending on the ecological system in question (e.g., Adamson 1927; Allan 1937; Ewel and 14

Bigelow 1996; Skottsberg 1929; Skottsberg 1913). The subdivisions of chamaephytes presented in 15

Ellenberg’s classification also encompass different life forms considering our classification (Table 16

1). Furthermore, they do not seem to enhance the system’s refinement (at least considering the case-17

study we presented here), since they showed correlation values (Mantel and procrustes analyses) 18

lower than those obtained with Raunkiaer’s system (Table 2). However, we do not intend to suggest 19

that these classic classifications should be abandoned. They have repeatedly proved to be applicable 20

and informative, and the decision to use them should be dependent on the ecological system and 21

specific questions. For example, Batalha and Martins (2002) suggested that Raunkiaer’s system is 22

not only applicable but also recommended considering the Brazilian Cerrado (tropical grasslands, or 23

Savannas), on the basis that local factors (fire, waterlogging, oligotrophism, extreme dry season and 24

aluminum toxicity) would be analogous to the unfavorable season his system was based upon. 25

Besides showing overall clearer vegetation patterns at a coarse level, our classification may 26

Page 71: abordagem quali-quantitativa e funcional de vegetação campestre ...

72

be used to more accurately describe between-site differences in vegetation, and to link these 1

differences to management history and local factors. Therophytes were strongly associated with the 2

most austral site (P1; Figure 3A, B), corroborating the hypothesized increase in importance of the 3

annual component in grassland vegetation towards higher latitudes. Differences between sites 4

regarding tussocks (solitary and connected), subshrubs and rhizomatous plants were clearly shown 5

by the averaged values of life form cover per biome (Figure 2A), and were largely responsible for 6

site segregation in the ordination analysis (Figure 3A). These differences are to a large extent the 7

reflection of land management. Atlantic Forest grasslands (sites A1, A2 and A3 in Figures 3 and 4) 8

are historically submitted to low stocking rates and yearly burning (Maraschin 2001; Nabinger et al. 9

2009), which favors tussock C4 species (Jacques 2003) and may hamper the establishment of 10

subshrubs and shrubs (Müller et al. 2007). Pampa sites, on the other hand, are managed with 11

moderate to high grazing pressures, which favors prostrate growth forms (Altesor et al. 2005; 12

Guerschman and Paruelo 2005). 13

The classification of life forms we presented and discussed here was conceived in and for 14

subtropical grasslands and perhaps to some degree to temperate grasslands. Its usefulness towards 15

higher temperate latitudes may be limited, since life forms we did not consider such as deciduous 16

shrubs may increase in importance (e.g., Campanella and Bertiller 2008). Also, it is likely that ‘true 17

hemicryptophytes’ (i.e., species that actually show partial or complete shoot reduction during part 18

of the year) should be found towards the tropics (tropical grasslands, or Savannas), in altitude 19

grasslands and in higher latitudes. Nevertheless, additional categories may be included in our 20

classification in such situations, just as categories that are too refined to a given situation may be 21

concatenated into a broader one (e.g., connected and isolated tussocks). Subtropical grasslands in 22

South America present a unique floristic composition, with coexistence of winter and summer 23

species, as well as some species also present in tropical and temperate areas (Boldrini 2009; Cabrera 24

and Willink 1980). Such species may present a variation in life form along their distribution, 25

behaving like evergreens in our case-study and like ‘true hemicriptophytes’ towards higher or lower 26

Page 72: abordagem quali-quantitativa e funcional de vegetação campestre ...

73

latitudes. This plasticity deserves further attention in future works. 1

We did not intend to create an all-encompassing classification applicable to different 2

ecosystems around the world, but our results have implications for researchers using plant life form 3

classifications elsewhere. Any life-form system applied needs to be appropriate to the ecological 4

system under study and must allow answering the questions asked. In our case, both Raunkiaer’s 5

and Ellenberg’s system proved not to distinguish different plant strategies in a way that reflected the 6

ecological properties and processes of the grassland system under study. Our suggestion of life form 7

classification proved to be a more accurate descriptor of subtropical grasslands and allowed to 8

perceive differences in functional plant composition between two subsets of our data that 9

corresponded to grasslands under somewhat different climate and under different management. 10

Generalizing these findings, we suggest that before using any existing classification, the relevance 11

of the underlying criteria to the ecosystem under study should be critically reflected, in order to get 12

meaningful answers. This seems to be important for other ecosystems worldwide, especially those 13

with extremely different evolutionary histories in comparison with the ecosystems in which classic 14

life form classifications were based on. Using easily obtained traits such as case-specific life form 15

sytems such as the one we propose here may be a promising starting point in initiatives to compare 16

communities in a regional scale, especially considering that a consistent trait database is still 17

lacking for these ecological systems. 18

While case-specific classifications can be successfully used to answer various ecological 19

questions in different ecosystems, they have one drawback: comparisons around the world may not 20

be possible. Here, working on a coarser scale and reflecting basically climatic differences, the 21

classical systems likely remain more useful. Nevertheless, we think that using a life form 22

classification consistently within ‘natural’ ecological units may reflect in future benefits, such as 23

facilitating and improving the accuracy of meta-analyses and allowing the development of unified 24

databases. This might be particularly important considering transnational biomes such as the 25

subtropical and temperate South American grasslands. 26

Page 73: abordagem quali-quantitativa e funcional de vegetação campestre ...

74

1

Acknowledgements 2

We would like to thank all members of the Grassland Vegetation Lab (LEVCamp), 3

especially the taxonomists, for discussions that have greatly improved this manuscript. The first 4

author received a scholarship by Coordenação de Aperfeiçoamento de Pessoal de Nível Superior 5

(CAPES). This study was part of the long-term ecological research project LTER/PELD Campos 6

Sulinos (CNPq 558282/2009-1). 7

Page 74: abordagem quali-quantitativa e funcional de vegetação campestre ...

75

Tables

Table 1. Hierarchical classification of plant life forms for subtropical and temperate grasslands,

with four levels of progressively refined categories and the respective acronym (A-D). R and E:

equivalent categories from Raunkiaer (1934) and Ellenberg and Mueller Dombois (1967)

classifications. Th = therophytes, G = geophytes, Ch = chamaephytes, Ph = phanerophytes, Gb =

bulbous geophytes, Gr = rhizomatous geophytes, HCh = herbaceous chamaephytes, RCh = reptant

herbaceous chamaephytes, SCh = sufrutescent chamaephytes, FCh = frutescent chamaephytes.

A B C D R E

Therophytes Therophytes Therophytes Therophytes Th Th Th

Geophytes Geophytes Geophytes Bulbous geophytes Bg G Gb

Rhizomatous geophytes Rg G Gr

Evergreens

Herbaceous

Prostrate

Prostate rosette evergreens Pr Ch HCh

Decumbent evergreens De Ch RCh

Rhizomatous evergreens Rh Ch HCh

Stoloniferous evergreens St Ch HCh

Erect

Solitary evergreen tussocks Te Ch HCh

Connected evergreen tussocks Ct Ch HCh

Evergreen forbs Ef Ch HCh

Erect rosette evergreens Er Ch HCh

Ligneous Ligneous Evergreen subshrubs Ss Ch SCh

Evergreen shrubs Sh Ph FCh

Table 2. Results from Mantel tests and Procrustes analyses evaluating the congruence between life-

form-weighted matrices and the original matrix of sampling areas described by species cover. SQ =

Sum of squares.

Mantel test Procrustes analysis

r statistic p-value SQ correlation p-value

New classification (D) 0.7594 0.0001 0.3123 0.8293 0.0001

New classification (C) 0.4868 0.0002 0.4706 0.7276 0.0001

Ellenberg & Muller-Dombois 0.2349 0.0078 0.7223 0.5270 0.0125

Raunkiaer 0.2605 0.0086 0.6992 0.5485 0.0044

Page 75: abordagem quali-quantitativa e funcional de vegetação campestre ...

76

Figures

Figure 1. Hierarchical classification of plant life forms for subtropical and temperate grasslands,

with four levels of progressively refined categories (A-D) and examples for each category (level D)

based on species from south Brazilian grasslands.

Page 76: abordagem quali-quantitativa e funcional de vegetação campestre ...

77

Figure 2. Cover values of life form categories from six grassland sites, averaged to biome level.

Life form classes according to the new classification (A), Raunkiaer (B) and Ellenberg & Mueller

Dombois (C). Th = therophytes, G = geophytes, Ch = chamaephytes, Ph = phanerophytes, Gb =

bulbous geophytes, Gr = rhizomatous geophytes, HCh = herbaceous chamaephytes, RCh = reptant

herbaceous chamaephytes, SCh = sufrutescent chamaephytes, FCh = frutescent chamaephytes. See

legends for life form categories from the new classification in Table 1.

Page 77: abordagem quali-quantitativa e funcional de vegetação campestre ...

78

Figure 3. Ordination diagrams of averaged cover values of life form categories from different

classification systems (A-D) as descriptors of sampling areas. Two levels of the proposed

classification are shown (‘D’ and ‘C’; see Table 1 for details). Legend: P = Pampa sites, A =

Atlantic Forest sites, Th = therophytes, G = geophytes, Ch = chamaephytes, Ph = phanerophytes,

Gb = bulbous geophytes, Gr = rhizomatous geophytes, HCh = herbaceous chamaephytes, RCh =

reptant herbaceous chamaephytes, SCh = sufrutescent chamaephytes, FCh = frutescent

chamaephytes. See legends for life form categories from the new classification in Table 1.

Page 78: abordagem quali-quantitativa e funcional de vegetação campestre ...

79

Figure 4. Procrustes analyses comparing ordinations of a species-described matrix of 18 grassland

sampling paddocks with matrices weighted by life form categories from different classification

systems. A. Ordinations of matrices weighted by life form categories. B. Procrustes rotation of

ordination axes, with original and target locations of sampling units. C. Procrustes errors. P = sites

from the Pampa biome; A = sites from the Atlantic Forest biome (also shaded in light gray).

Page 79: abordagem quali-quantitativa e funcional de vegetação campestre ...

80

electronic supplementary material

ESM 1 - Life form categories for the 193 grassland plant species used in the analyses. te = solitary

evergreen tussocks; st = stolonoiferous evergreens; rh = rhizomatous evergreens; ss = evergreen

subshrubs; de = decumbent evergreens; ef = evergreen forbs; ct = connected evergreen tussocks; pr

= prostate rosette evergreens; er = erect rosette evergreens; th = therophytes; bg = bulbous

geophytes; sh = evergreen shrubs

Family Species Author Life form

Acanthaceae Stenandrium diphyllum Nees pr

Amaranthaceae Pfaffia tuberosa Hicken ef

Apiaceae Cyclospermum leptophyllum (Pers.) Sprague ex Britton & P. Wilson th

Apiaceae Eryngium ebracteatum Lam. pr

Apiaceae Eryngium echinatum Urb. er

Apiaceae Eryngium horridum Malme er

Apiaceae Eryngium nudicaule Lam. pr

Araliaceae Hydrocotyle exigua Malme st

Asteraceae Achyrocline satureioides (Lam.) DC. ef

Asteraceae Acmella bellidioides (Smith in Rees) R.K. Jansen ef

Asteraceae Aster squamatus (Spreng.) Hieron. ss

Asteraceae Baccharis coridifolia DC. ss

Asteraceae Baccharis dracunculifolia DC. sh

Asteraceae Baccharis genistelloides (Lam.) Pers. ss

Asteraceae Baccharis pentodonta Malme ss

Asteraceae Baccharis riograndensis I.L. Teodoro & J.E. Vidal ss

Asteraceae Baccharis subtropicalis G. Heiden ss

Asteraceae Baccharis tridentata Vahl ss

Asteraceae Calyptocarpus biaristatus (DC.) H. Rob. th

Asteraceae Chaptalia exscapa (Pers.) Baker pr

Asteraceae Chaptalia integerrima (Vell.) Burkart pr

Asteraceae Chaptalia piloselloides (Vahl) Baker pr

Asteraceae Chaptalia runcinata Kunth pr

Asteraceae Chevreulia acuminata Less. de

Asteraceae Chevreulia revoluta A.A. Schneid. & R. Trevis. ef

Asteraceae Chevreulia sarmentosa (Pers.) S.F. Blake st

Asteraceae Conyza bonariensis (L.) Cronquist th

Asteraceae Elephantopus mollis Kunth ef

Asteraceae Erigeron primulifolium (Lam.) Greuter th

Asteraceae Eupatorium squarrulosum Hook. & Arn. ef

Asteraceae Gamochaeta americana (Mill.) Wedd. ef

Asteraceae Hypochaeris catharinensis Cabrera pr

Asteraceae Lucilia linearifolia Baker ef

Page 80: abordagem quali-quantitativa e funcional de vegetação campestre ...

81

Family Species Author Life form

Asteraceae Noticastrum decumbens (Baker) Cuatrec. de

Asteraceae Panphalea araucariophila Cabrera th

Asteraceae Panphalea heterophylla Less. th

Asteraceae Pterocaulon virgatum (L.) DC. ef

Asteraceae Senecio heterotrichius DC. ss

Asteraceae Senecio madagascariensis Poir. th

Asteraceae Soliva sessilis Ruiz & Pav. th

Asteraceae Trichocline catharinensis Cabrera pr

Campanulaceae Wahlenbergia linarioides (Lam.) A. DC. ef

Caryophyllaceae Cerastium glomeratum Thuill. th

Caryophyllaceae Spergularia sp. - th

Convolvulaceae Dichondra macrocalyx Meisn. st

Convolvulaceae Dichondra sericea Sw. st

Convolvulaceae Evolvulus sericeus Sw. de

Cyperaceae Bulbostylis capillaris (L.) C.B. Clarke te

Cyperaceae Bulbostylis juncoides (Vahl) Kük. ex Osten te

Cyperaceae Bulbostylis sp. Kunth te

Cyperaceae Bulbostylis sphaerocephala (Boeckeler) C.B. Clarke te

Cyperaceae Carex bonariensis Desf. ex Poir. te

Cyperaceae Carex phalaroides Kunth ct

Cyperaceae Carex sororia Kunth te

Cyperaceae Cyperaceae sp. - te

Cyperaceae Cyperus aggregatus (Willd.) Endl. te

Cyperaceae Cyperus hermaphroditus (Jacq.) Standl. te

Cyperaceae Cyperus reflexus Vahl te

Cyperaceae Eleocharis dunensis Kük. ct

Cyperaceae Eleocharis nudipes (Kunth) H. Pfeiff. te

Cyperaceae Eleocharis viridans Kük. ex Osten ct

Cyperaceae Fimbristylis dichotoma (L.) Vahl te

Cyperaceae Kyllinga odorata Vahl te

Cyperaceae Kyllinga vaginata Lam. ct

Cyperaceae Lipocarpha humboldtiana Nees te

Cyperaceae Rhynchospora barrosiana Guagl. te

Cyperaceae Rhynchospora emaciata (Nees) Boeckeler ct

Cyperaceae Rhynchospora flexuosa C.B. Clarke te

Cyperaceae Rhynchospora megapotamica (Spreng.) H. Pfeiff. st

Cyperaceae Rhynchospora sp. - te

Cyperaceae Rhynchospora tenuis Willd. ex Link ct

Cyperaceae Scleria distans Poir. ct

Euphorbiaceae Euphorbia selloi (Klotzsch & Garcke) Boiss. de

Page 81: abordagem quali-quantitativa e funcional de vegetação campestre ...

82

Family Species Author Life form

Fabaceae Aeschynomene falcata (Poir.) DC. de

Fabaceae Crotalaria hilariana Benth. ef

Fabaceae Desmanthus tatuhyensis Hoehne ss

Fabaceae Desmanthus virgatus (L.) Willd. ss

Fabaceae Desmodium incanum (Sw.) DC. st

Fabaceae Galactia marginalis Benth. ss

Fabaceae Galactia neesii DC. ss

Fabaceae Macroptilium gibbosifolium (Ortega) A. Delgado st

Fabaceae Macroptilium prostratum (Benth.) Urb. de

Fabaceae Rhynchosia corylifolia Mart. ex Benth. de

Fabaceae Stylosanthes montevidensis Vogel ef

Fabaceae Trifolium polymorphum Poir. st

Fabaceae Trifolium riograndense Burkart st

Hypoxidaceae Hypoxis decumbens L. bg

Iridaceae Herbertia lahue (Molina) Goldblatt bg

Iridaceae Sisyrinchium micranthum Cav. th

Iridaceae Sisyrinchium palmifolium L. ef

Iridaceae Sisyrinchium platense I.M. Johnst. ef

Juncaceae Juncus capillaceus Lam. te

Juncaceae Juncus dichotomus Elliott te

Juncaceae Juncus microcephalus Kunth te

Juncaceae Juncus tenuis Willd. te

Lamiaceae Cunila galioides Benth. ss

Lamiaceae Scutellaria racemosa Pers. ef

Malvaceae Ayenia mansfeldiana (Herter) Herter ex Cristóbal ef

Melastomataceae Acisanthera alsinaefolia (DC.) Triana ef

Melastomataceae Tibouchina gracilis (Bonpl.) Cogn. ef

Myrtaceae Campomanesia aurea O. Berg sh

Orobanchaceae Agalinis communis (Cham. & Schltdl.) D'Arcy th

Oxalidaceae Oxalis brasiliensis G. Lodd. bg

Oxalidaceae Oxalis eriocarpa DC. st

Oxalidaceae Oxalis lasiopetala Zucc. bg

Plantaginaceae Plantago australis Lam. pr

Plantaginaceae Plantago myosuros Lam. th

Plantaginaceae Plantago penantha Griseb. th

Plantaginaceae Scoparia dulcis L. ss

Poaceae Agrostis hygrometrica Nees th

Poaceae Agrostis montevidensis Spreng. ex Nees th

Poaceae Andropogon lateralis Nees te

Poaceae Andropogon macrothrix Trin. te

Page 82: abordagem quali-quantitativa e funcional de vegetação campestre ...

83

Family Species Author Life form

Poaceae Andropogon ternatus (Spreng.) Nees te

Poaceae Aristida flaccida Trin. & Rupr. te

Poaceae Aristida murina Cav. te

Poaceae Aristida uruguayensis Henrard te

Poaceae Aristida venustula Arechav. te

Poaceae Axonopus affinis Chase st

Poaceae Axonopus argentinus Parodi te

Poaceae Axonopus compressus (Sw.) P. Beauv. st

Poaceae Axonopus fissifolius (Raddi) Kuhlm. st

Poaceae Axonopus siccus (Nees) Kuhlm. te

Poaceae Axonopus suffultus (Mikan ex Trin.) Parodi te

Poaceae Bothriochloa laguroides (DC.) Herter te

Poaceae Briza minor L. th

Poaceae Briza uniolae (Nees) Nees ex Steud. te

Poaceae Chascolytrum poomorphum (J. Presl) L. Essi, Longhi-Wagner & Souza-Chies te

Poaceae Chascolytrum subaristatum (Lam.) Desv. te

Poaceae Chloris grandiflora Roseng. & Izag. te

Poaceae Cynodon dactylon (L.) Pers. st

Poaceae Danthonia cirrata Hack. & Arechav. te

Poaceae Danthonia secundiflora J. Presl te

Poaceae Dichanthelium sabulorum (Lam.) Gould & C.A. Clark de

Poaceae Eragrostis airoides Nees te

Poaceae Eragrostis neesii Trin. th

Poaceae Eragrostis polytricha Nees te

Poaceae Eustachys brevipila (Roseng. & Izag.) Caro & E.A. Sánchez te

Poaceae Melica eremophila Torres te

Poaceae Melica rigida Cav. te

Poaceae Microchloa indica (L. f.) P. Beauv. te

Poaceae Mnesithea selloana (Hack.) de Koning & Sosef te

Poaceae Paspalum almum Chase te

Poaceae Paspalum compressifolium Swallen te

Poaceae Paspalum dilatatum Poir. te

Poaceae Paspalum lepton Schult. rh

Poaceae Paspalum maculosum Trin. te

Poaceae Paspalum notatum Alain ex Flüggé rh

Poaceae Paspalum plicatulum Michx. te

Poaceae Paspalum polyphyllum Nees ex Trin. te

Poaceae Paspalum pumilum Nees rh

Poaceae Piptochaetium lasianthum Griseb. te

Poaceae Piptochaetium montevidense (Spreng.) Parodi te

Page 83: abordagem quali-quantitativa e funcional de vegetação campestre ...

84

Family Species Author Life form

Poaceae Piptochaetium stipoides (Trin. & Rupr.) Hack. ex Arechav. te

Poaceae Saccharum angustifolium (Nees) Trin. te

Poaceae Saccharum sp. - te

Poaceae Sacciolepis vilvoides (Trin.) Chase te

Poaceae Schizachyrium condensatum (Kunth) Nees te

Poaceae Schizachyrium hatschbachii Peichoto te

Poaceae Schizachyrium microstachyum (Desv. ex Ham.) Roseng., B.R. Arrill. & Izag. te

Poaceae Schizachyrium spicatum (Spreng.) Herter te

Poaceae Schizachyrium tenerum Nees te

Poaceae Setaria fiebrigii R.A.W. Herrm. te

Poaceae Setaria parviflora (Poir.) Kerguélen te

Poaceae Sorghastrum sp. - te

Poaceae Sporobolus indicus (L.) R. Br. te

Poaceae Sporobolus monandrus Roseng., B.R. Arrill. & Izag. te

Poaceae Steinchisma decipiens (Nees ex Trin.) W.V. Br. te

Poaceae Steinchisma hians (Elliott) Nash te

Poaceae Stipa filiculmis Delile te

Poaceae Stipa juergensii Hack. te

Poaceae Stipa setigera J. Presl te

Poaceae Stipa sp. - te

Poaceae Stipa tenuiculmis Hack. te

Poaceae Trachypogon montufari (Kunth) Nees te

Poaceae Trachypogon montufarii var. mollis (Nees) Andersson te

Poaceae Vulpia bromoides (L.) Gray th

Polygalaceae Polygala linoides Poir. th

Polygalaceae Polygala molluginifolia A. St.-Hil. & Moq. ef

Rosaceae Acaena eupatoria Cham. & Schltdl. de

Rubiaceae Borreria eryngioides Cham. & Schltdl. ef

Rubiaceae Galium humile Cham. & Schltdl. de

Rubiaceae Galium richardianum (Gillies ex Hook. & Arn.) Endl. ex Walp. de

Rubiaceae Richardia humistrata (Cham. & Schltdl.) Steud. st

Rubiaceae Richardia stellaris (Cham. & Schltdl.) Steud. de

Rubiaceae Spermacoce verticillata L. ef

Selaginellaceae Selaginella sp. - de

Solanaceae Nierembergia riograndensis Hunz. & A.A. Cocucci ss

Solanaceae Solanum atropurpureum Schrank ss

Verbenaceae Glandularia selloi (Spreng.) Tronc. st

Verbenaceae Lippia angustifolia Cham. ss

Verbenaceae Phyla nodiflora (L.) Greene st

Verbenaceae Verbena montevidensis Spreng. ss

Page 84: abordagem quali-quantitativa e funcional de vegetação campestre ...

85

References

Adamson R. (1939) The classification of life-forms of plants. The Botanical Review 5, 546-61.

Adamson R. S. (1927) The plant communities of Table Mountain: preliminary account. Journal

of Ecology 15, 278-309.

Allan H. H. (1937) A consideration of the" biological spectra" of New Zealand. The Journal of

Ecology, 116-52.

Altesor A., Oesterheld M., Leoni E., Lezama F. & Rodriguez C. (2005) Effect of grazing on

community structure and productivity of a Uruguayan grassland. Plant Ecology 179, 83-91.

Altesor A., Piñeiro G., Lezama F., Jackson R., Sarasola M. & Paruelo J. (2006) Ecosystem

changes associated with grazing in subhumid South American grasslands. Journal of Vegetation

Science 17, 323-32.

Areschoug F. V. K. (1896) Beiträge zur Biologie der geophilen Pflanzen. Malmström, Lund.

Arnold J. F. (1955) Plant life-form classification and its use in evaluating range conditions and

trend. Journal of Range Management 8, 176-81.

Aronson M. F., Handel S. N. & Clemants S. E. (2007) Fruit type, life form and origin determine

the success of woody plant invaders in an urban landscape. Biological Invasions 9, 465-75.

Batalha M. & Martins F. (2002) Biological spectra of cerrado sites. Flora 197, 452-60.

Behling H., Pillar V. D. & Bauermann S. G. (2005) Late Quaternary grassland (Campos), gallery

forest, fire and climate dynamics, studied by pollen, charcoal and multivariate analysis of the São

Francisco de Assis core in western Rio Grande do Sul (southern Brazil). Review of Palaeobotany

and Palynology 133, 235-48.

Behling H., Pillar V. D. P., Orlóci L. & Bauermann S. G. (2004) Late Quaternary Araucaria

forest, grassland (Campos), fire and climate dynamics, studied by high-resolution pollen, charcoal

and multivariate analysis of the Cambará do Sul core in southern Brazil. Palaeogeography,

Palaeoclimatology, Palaeoecology 203, 277-97.

Page 85: abordagem quali-quantitativa e funcional de vegetação campestre ...

86

Bhattarai K. R. & Vetaas O. R. (2003) Variation in plant species richness of different life forms

along a subtropical elevation gradient in the Himalayas, east Nepal. Global Ecology and

Biogeography 12, 327-40.

Bilenca D. & Miñarro F. (2004) Identificación de áreas valiosas de pastizal en las pampas y

campos de Argentina, Uruguay y sur de Brasil. Fundación Vida Silvestre Argentina, Buenos Aires.

Boldrini I. I. (2009) A flora dos campos do Rio Grande do Sul. Campos sulinos: conservação e

uso sustentável da biodiversidade. Brasília: Ministério do Meio Ambiente, 63-77.

Boyle B., Hopkins N., Lu Z., Garay J. A. R., Mozzherin D., Rees T., Matasci N., Narro M. L.,

Piel W. H. & Mckay S. J. (2013) The taxonomic name resolution service: an online tool for

automated standardization of plant names. BMC bioinformatics 14, 16.

Braun-Blanquet J. (1928) Pflanzensoziologie. Grundzüge der Vegetationskunde. Biologische

Studienbücher 7.

Bredenkamp G., Spada F. & Kazmierczak E. (2002) On the origin of northern and southern

hemisphere grasslands. Plant Ecology 163, 209-29.

Burkart A. (1975) Evolution of grasses and grasslands in South America. Taxon, 53-66.

Cabrera A. L. & Willink A. (1980) Biogeografía de América latina. Serie de Biología.

Monografías 13.

Campanella M. & Bertiller M. B. (2008) Plant phenology, leaf traits and leaf litterfall of

contrasting life forms in the arid Patagonian Monte, Argentina. Journal of Vegetation Science 19,

75-85.

Castanho C. T., Oliveira A. A. & Prado P. I. (2012) The importance of plant life form on spatial

associations along a subtropical coastal dune gradient. Journal of Vegetation Science 23, 952-61.

Clements F. E. (1920) Plant indicators: the relation of plant communities to process and practice.

Carnegie Institution of Washington.

Collins S. L. & Calabrese L. B. (2012) Effects of fire, grazing and topographic variation on

vegetation structure in tallgrass prairie. Journal of Vegetation Science 23, 563-75.

Page 86: abordagem quali-quantitativa e funcional de vegetação campestre ...

87

Cornelissen J., Lavorel S., Garnier E., Diaz S., Buchmann N., Gurvich D., Reich P., Ter Steege

H., Morgan H. & Van Der Heijden M. (2003) A handbook of protocols for standardised and easy

measurement of plant functional traits worldwide. Australian Journal of Botany 51, 335-80.

Cramer M. D., Wakeling J. L. & Bond W. J. (2012) Belowground competitive suppression of

seedling growth by grass in an African savanna. Plant Ecology 213, 1655-66.

Darwin C. (1859) On the origins of species by means of natural selection. London: Murray.

De Candolle A. P. (1818) Regni Vegetabolis Systema Naturale, Paris.

Díaz S. & Cabido M. (1997) Plant functional types and ecosystem function in relation to global

change. Journal of vegetation science 8, 463-74.

Diaz S., Lavorel S., McIntyre S., Falczuk V., Casanoves F., Milchunas D. G., Skarpe C., Rusch

G., Sternberg M. & NOY‐MEIR I. (2007) Plant trait responses to grazing–a global synthesis.

Global Change Biology 13, 313-41.

Drude O. (1896) Deutschlands Pflanzengeographie: ein geographisches Charakterbild der Flora

von Deutschland und den angrenzenden Alpen- sowie Karpathenländern. J. Engelhorn.

Du Rietz G. E. (1921) Zur methodologischen Grundlage der modernen Pflanzensoziologie.

Upsala.

Du Rietz G. E. (1931) Life-forms of terrestrial flowering plants. Almqvist & Wiksell.

Ellenberg D. & Mueller-Dombois D. (1974) Aims and methods of vegetation ecology. Wiley

New York, NY.

Ellenberg H. & Mueller Dombois D. (1967) A key to Raunkiaer plant life forms with revised

subdivisions. Ber. geobot. Inst. eidg. tech. Hochschule Rubel 37, 56-73.

Ewel J. J. & Bigelow S. W. (1996) Plant life-forms and tropical ecosystem functioning. In:

Biodiversity and ecosystem processes in tropical forests pp. 101-26. Springer.

Fidelis A., Appezzato-da-Glória B., Pillar V. D. & Pfadenhauer J. (2014) Does disturbance

affect bud bank size and belowground structures diversity in Brazilian subtropical grasslands?

Flora-Morphology, Distribution, Functional Ecology of Plants 209(2):110–116.

Page 87: abordagem quali-quantitativa e funcional de vegetação campestre ...

88

Fidelis A. T., Delgado Cartay M. D., Blanco C. C., Muller S. C., Pillar V. d. P. & Pfadenhauer J.

S. (2010) Fire intensity and severity in Brazilian Campos grasslands. Interciencia: revista de ciencia

y tecnologia de america. Caracas. Vol. 35, n. 10, p. 739-745.

Garcia E. N., Boldrini I. I. & Jacques A. V. A. (2002) Dinâmica de formas vitais de uma

vegetação campestre sob diferentes práticas de manejo e exclusão. Iheringia. Série botânica 57,

215-41.

Gómez-Aparicio L. (2009) The role of plant interactions in the restoration of degraded

ecosystems: a meta‐analysis across life‐forms and ecosystems. Journal of Ecology 97, 1202-14.

Grisebach A. (1872) Die Vegetation der Erde nach ihrer klimatischen Anordnung. Leipzig.

Guerschman J. P. & Paruelo J. M. (2005) Agricultural impacts on ecosystem functioning in

temperate areas of North and South America. Global and Planetary Change 47, 170-80.

Hadar L., Noy‐Meir I. & Perevolotsky A. (1999) The effect of shrub clearing and grazing on the

composition of a Mediterranean plant community: functional groups versus species. Journal of

Vegetation Science 10, 673-82.

Hort A. (1916) Enquiry into plants and minor works on odours and weather signs. William

Heinemann, London.

Huang Q. Q., Wu J. M., Bai Y. Y., Zhou L. & Wang G. X. (2009) Identifying the most noxious

invasive plants in China: role of geographical origin, life form and means of introduction.

Biodiversity and conservation 18, 305-16.

Humboldt A. V. (1806) Ideen zu einer Physiognomik der Gewächse. Tübingen.

Ivanova L. (2012) Restructuring of the leaf mesophyll in a series of plant life forms. In: Doklady

Biological Sciences.

Jacques A. V. A. (2003) A queima das pastagens naturais–efeitos sobre o solo ea vegetação.

Ciência Rural 33, 177-81.

Kerner A. (1869) Die Abhängigkeit der Pflanzengestalt von Klima und Boden. Festschriften der

Versammlungen Deutscher Naturforscher und Ärzte, Bd 6, 29-45.

Page 88: abordagem quali-quantitativa e funcional de vegetação campestre ...

89

Kerner A. (1929) Das Pflanzenleben der Donauländer. Universitäts-Verlag Wagner.

Lezama F., Altesor A., León R. J. & Paruelo J. M. (2006) Heterogeneidad de la vegetación en

pastizales naturales de la región basáltica de Uruguay. Ecología Austral 16, 167-82.

Lezama F., Baeza S., Altesor A., Cesa A., Chaneton E. J. & Paruelo J. M. (2013) Variation of

grazing‐induced vegetation changes across a large‐scale productivity gradient. Journal of

Vegetation Science.

López R. P., Valdivia S., Rivera M. L. & Rios R. S. (2013) Co-occurrence Patterns along a

Regional Aridity Gradient of the Subtropical Andes Do Not Support Stress Gradient Hypotheses.

PloS one 8, e58518.

Maraschin G. E. (2001) Production potential of South American grasslands. In: XIX

International Grassland Congress, São Paulo, Brazil.

McIntyre S., Lavorel S. & Tremont R. (1995) Plant life-history attributes: their relationship to

disturbance response in herbaceous vegetation. Journal of Ecology, 31-44.

Milchunas D., Sala O. & Lauenroth W. K. (1988) A generalized model of the effects of grazing

by large herbivores on grassland community structure. American Naturalist, 87-106.

Moustakas A., Kunin W. E., Cameron T. C. & Sankaran M. (2013) Facilitation or competition?

Tree effects on grass biomass across a precipitation gradient. PloS one 8, e57025.

Müller S. C., Overbeck G. E., Pfadenhauer J. & Pillar V. D. (2007) Plant functional types of

woody species related to fire disturbance in forest–grassland ecotones. Plant Ecology 189, 1-14.

Nabinger C., Ferreira E., Freitas A., Carvalho P. & Sant’Anna D. (2009) Produção animal com

base no campo nativo: aplicações de resultados de pesquisa. Campos sulinos: conservação e uso

sustentável da biodiversidade. Brasília: Ministério do Meio Ambiente, 175-98.

Nelis L. C. (2012) Life Form and Life History Explain Variation in Population Processes in a

Grassland Community Invaded by Exotic Plants and Mammals. PloS one 7, e42906.

Oksanen J., Blanchet F., Kindt R., Legendre P., O'Hara R., Simpson G., Solymos P., Stevens M.

& Wagner H. (2013) vegan: Community Ecology Package. 2013. R package version 2.0-8.

Page 89: abordagem quali-quantitativa e funcional de vegetação campestre ...

90

Overbeck G. E., Müller S., Pillar V. & Pfadenhauer J. (2006) Floristic composition,

environmental variation and species distribution patterns in burned grassland in southern Brazil.

Brazilian Journal of Biology 66, 1073-90.

Overbeck G. E., Müller S. C., Pillar V. D. & Pfadenhauer J. (2005) Fine‐scale post‐fire

dynamics in southern Brazilian subtropical grassland. Journal of Vegetation Science 16, 655-64.

Overbeck G. E. & Pfadenhauer J. (2007) Adaptive strategies in burned subtropical grassland in

southern Brazil. Flora-Morphology, Distribution, Functional Ecology of Plants 202, 27-49.

Pekin B. K., Wittkuhn R. S., Boer M. M., Macfarlane C. & Grierson P. F. (2012) Response of

plant species and life form diversity to variable fire histories and biomass in the jarrah forest of

south‐west Australia. Austral Ecology 37, 330-8.

Peres-Neto P. R. & Jackson D. A. (2001) How well do multivariate data sets match? The

advantages of a Procrustean superimposition approach over the Mantel test. Oecologia 129, 169-78.

Pérez-Harguindeguy N., Díaz S., Garnier E., Lavorel S., Poorter H., Jaureguiberry P., Bret-Harte

M., Cornwell W., Craine J. & Gurvich D. (2013) New handbook for standardised measurement of

plant functional traits worldwide. Australian Journal of Botany.

Pillar V. (1997) Multivariate exploratory analysis and randomization testing with MULTIV.

Coenoses 12, 145-8.

Pillar V. D. & Duarte L. d. S. (2010) A framework for metacommunity analysis of phylogenetic

structure. Ecology letters 13, 587-96.

Pillar V. D., Duarte L. d. S., Sosinski E. E. & Joner F. (2009) Discriminating trait‐convergence

and trait‐divergence assembly patterns in ecological community gradients. Journal of Vegetation

Science 20, 334-48.

Pillar V. D. P. & Orlóci L. (1993) Taxonomy and perception in vegetation analysis. Coenosis 8,

53-66.

Pillar V. D. P. & Orlóci L. (1996) On randomization testing in vegetation science: multifactor

comparisons of relevé groups. Journal of Vegetation Science 7, 585-92.

Page 90: abordagem quali-quantitativa e funcional de vegetação campestre ...

91

Pillar V. D. P. & Quadros F. L. F. (1997) Grassland-forest boundaries in southern Brazil.

Coenoses 12, 119-26.

R Development Core Team. (2012) R: A Language and Environment for Statistical Computing.

R Foundation for Statistical Computing, Vienna, Austria.

Raunkiaer C. (1904) Om biologiske Typer, med Hensyn til Planternes Tilpasning til at overleve

ugunstige Aarstider. Botanisk Tidsskrift 26.

Raunkiaer C. (1905) Types biologiques pour la géographie botanique. Danske Videnskabernes

Selskabs Forhandl 5, 347-437.

Raunkiaer C. (1907) Planterigets Livsformer og deres Betydning for Geografien. Kobenhaven.

See also the English translation in" The life forms of plants and statistical plant geography," 1934.

1908. Livsformernes statistik som grundlag for biologisk plantegeografi. Bot. Tidssk 29, 42-83.

Raunkiaer C. (1934) The life forms of plants and statistical plant geography. Claredon, Oxford.

Reiter H. (1885) Die consolidation der Physiognomik: Als versuch einer Oekologie der

Gewaechse. Leuschner & Lubensky.

Sarmiento G. & Monasterio M. (1983) Life forms and phenology. Ecosystems of the world 13,

79-108.

Skaer M. J., Graydon D. J. & Cushman J. (2012) Community‐level consequences of cattle

grazing for an invaded grassland: variable responses of native and exotic vegetation. Journal of

Vegetation Science.

Skottsberg C. (1929) Plant Communities of the Juan Fernandez Islands. In: International

congress of Plant Sciences, Ithaca, New York.

Skottsberg C. J. (1913) A botanical survey of the Falkland Islands. Almqvist & Wiksells Boktr.

Solbrig O. (1993) Plant traits and adaptive strategies: their role in ecosystem function. In:

Biodiversity and ecosystem function pp. 97-116. Springer.

Soriano A., León R., Sala O., Lavado R., Deregibus V., Cauhépé M., Scaglia O., Velázquez C.

& Lemcoff J. (1992) Río de la Plata Grasslands. In ‘Ecosystems of the world 8A. Natural

Page 91: abordagem quali-quantitativa e funcional de vegetação campestre ...

92

grasslands. Introduction and Western Hemisphere’.(Ed. RT Coupland) pp. 367–407. Elsevier:

Amsterdam.

Sylvén N. (1906) Om de svenska dikotyledonernas första förstärkningsstadium. Kungliga

Svenska Vetenskapsakademiens Handlingar 40, 1-348.

Tsujino R. & Yumoto T. (2013) Vascular plant species richness along environmental gradients

in a cool temperate to sub-alpine mountainous zone in central Japan. Journal of plant research, 1-12.

Vashistha R., Rawat N., Chaturvedi A., Nautiyal B., Prasad P. & Nautiyal M. (2011)

Characteristics of life-form and growth-form of plant species in an alpine ecosystem of North-West

Himalaya. Journal of Forestry Research 22, 501-6.

Vesque J. (1882) L'èspèce végétale considérée au point de vue de l'anatomie comparée. Ann.

Sci. Nat. Bot 6.

Vogel S. & Ewel K. C. (1972) A Model Menagerie: Laboratory Studies About Living Systems.

Addison-Wesley Publishing Company.

Walker B. H. (1992) Biodiversity and ecological redundancy. Conservation Biology 6, 18-23.

Warming E. (1895) Plantesamfund: grundtræk af den økologiske plantegeografi. Philipsen.

Warming E. & Vahl M. (1909) Oecology of plants: an introduction to the study of plant

communities. Clarendon. Oxford.

Weiher E., Werf A., Thompson K., Roderick M., Garnier E. & Eriksson O. (1999) Challenging

Theophrastus: a common core list of plant traits for functional ecology. Journal of Vegetation

Science 10, 609-20.

Westoby M. (1998) A leaf-height-seed (LHS) plant ecology strategy scheme. Plant and soil 199,

213-27.

Westoby M. (2006) Phylogenetic ecology at world scale, a new fusion between ecology and

evolution. Ecology 87, S163-S5.

Whittaker R. (1975) Community and ecosystems. McMillan, New York.

Page 92: abordagem quali-quantitativa e funcional de vegetação campestre ...

93

Willems J. H. (1985) Growth form spectra and species diversity in permanent grassland plots

with different management. Sukzession auf Grünlandbrachen, 35-43.

Page 93: abordagem quali-quantitativa e funcional de vegetação campestre ...

94

Page 94: abordagem quali-quantitativa e funcional de vegetação campestre ...

95

Exploring the relationships between habitat structure and arthropod diversity in South

Brazilian grasslands: a functional perspective

Pedro M.A. Ferreira, Luciana R. Podgaiski, Gerhard E. Overbeck, Ilsi I. Boldrini

P.M.A. Ferreira (corresponding author)

Universidade Federal do Rio Grande do Sul, Programa de Pós Graduação em Botânica, Av. Bento

Gonçalves 9500 Bloco IV, P. 43432, CEP 91501-970, Porto Alegre, RS, Brazil.

e-mail: [email protected]

phone: +55 51 3308 7555

Page 95: abordagem quali-quantitativa e funcional de vegetação campestre ...

96

Abstract

Habitat structure is one of the major factors shaping biodiversity patterns, and plays a central role in

ecological theory. Plant species define habitat structure in most ecosystems, and structurally

complex habitats provide more niches and different ways of using resources, ultimately increasing

diversity at multiple taxonomic levels. In grasslands, habitat structure is strongly influenced by

disturbance regimes of fire and grazing. Diversity patterns of grassland arthropods are determined

by plant diversity and structural architecture, although empirical data supporting this relationhip is

limited for South American subtropical ecosystems. In this paper we explored the relationships

between plant richness and diversity, habitat structure, and arthropod communities in six grassland

sites under different grazing pressures from southern Brazil. We used four major groups of

arthropod (Coleoptera, Araneae, Hemiptera and Formicidae) and vegetation height and plant

functional diversity based on life form traits as descriptors of habitat structure. We searched for

relationships between habitat structure and (1) plant species richness and diversity and (2) arthropod

richness and diversity. Also, we explored the relationship between functional traits of arthropods

and plants and estimate the magnitude of co-variation between community-weighted mean traits

(CWM) using co-inertia analysis. Vegetation vertical structure and grazing pressure were, as

expected, correlated. Sites under lower grazing pressure showed decreased habitat heterogeneity

and plant richness and diversity. Plant diversity indexes and vegetation height were negatively

correlated. Richness of arthropod orders was positively correlated to plant diversity and functional

diversity. Co-inertia analyses showed a significant association between matrices with CWM traits

of plants and all arthropod groups. The association was stronger in Coleoptera and Araneae and

weaker in Hemiptera and Formicidae. Differences in patterns of association are related to different

relationships between habitat structure and arthropod community in each group. Also, we found

strong and group-specific pairwise correlations between arthropod and plant traits, providing further

evidence that each group of arthropod responds differently to vegetation structure. Our results

provided further evidence supporting the close relationship between grazing and grassland

vegetation structure and diversity patterns. Grazing promotes habitat heterogeneity, which is turn

influences diversity patterns of arthropods. Our results may guide future research focused on

specific arthropod groups in subtropical grasslands.

Keywords: subtropical grasslands, plant life forms, co-inertia analysis, land management,

disturbance, grazing

Page 96: abordagem quali-quantitativa e funcional de vegetação campestre ...

97

Introduction 1

2

Habitat structure is one of the factors that define biodiversity patterns in most ecosystems, 3

and plays a central role in ecological theory (e.g., MacArthur, R. H. 1967, Simpson, E. H. 1949). 4

According to the habitat heterogeneity hypothesis, habitats with complex structures provide more 5

niches and different ways of using resources, ultimately increasing diversity (Bazzaz, F. 1975). 6

Plant communities determine habitat structure in most terrestrial ecosystems, and as a result affect 7

patterns of distribution and interaction of animal species (McCoy, E. and Bell, S. 1991, Tews, J. et 8

al. 2004). Although a positive correlation between habitat heterogeneity and animal species 9

diversity has been repeatedly found in many ecosystems and for different taxa, actual empirical 10

support for this relationship is strongly biased towards vertebrate groups and non-natural habitats 11

(review in Tews, J. et al. 2004). Moreover, the influence of habitat structure on animals is mostly 12

evaluated on the basis of animal species richness and diversity, whereas functional aspects are often 13

ignored (but see Podgaiski, L. R. et al. 2013). Understanding trait-environment links is a key step to 14

determine community assembly in ecosystems, which has been shown to be important in studies on 15

the animal response to habitat change after disturbances (Vandewalle, M. et al. 2010). 16

In grasslands, habitat structure is strongly influenced by disturbance regimes of fire and 17

grazing, which in fact are the evolutionary drivers and maintainers of these ecosystems (Behling, H. 18

et al. 2004, Bond, W. J. and Keeley, J. E. 2005, Knapp, A. K. et al. 1998, Milchunas, D. et al. 19

1988). Despite the apparent structural simplicity at first glance, grasslands can show high habitat 20

heterogeneity, arising not only from disturbance (Fuhlendorf, S. D. and Engle, D. M. 2001, Kruess, 21

A. and Tscharntke, T. 2002, Wardle, D. A. et al. 2005), but also because they can show a vast array 22

of life forms and high species richness. Although grasses are responsible for ca. 80% of the 23

aboveground biomass of grasslands, forbs and other life forms make up ca. 80% of the total plant 24

species richness in these systems (Knapp, A. K. et al. 1998). Fire and grazing promote habitat 25

heterogeneity by creating patches under different levels of disturbance and consequently selecting 26

Page 97: abordagem quali-quantitativa e funcional de vegetação campestre ...

98

plant species that share similar traits (including structural traits that can are considered in life form 1

classifications) compatible with each condition (Bond, W. et al. 2005, Diaz, S. et al. 2007, Grime, J. 2

P. 2006). Although the heterogeneity promoted by filtering mechanisms is scale-dependent, any 3

change in plant community structure will also have consequences for the diversity of coexisting 4

animal species, which in turn may feedback to ecosystem function (Loreau, M. et al. 2002, Tews, J. 5

et al. 2004). 6

Especially patterns of arthropod diversity in grassland ecosystem are, to a large extent, 7

dependent on the plant diversity and structural architecture (review in Joern, A. and Laws, A. N. 8

2013). Arthropods comprise a large portion of grassland diversity (Fay, P. A. 2003, Tscharntke, T. 9

and Greiler, H.-J. 1995) and also play key roles on ecosystem functioning (Belovsky, G. and Slade, 10

J. 2000, Meyer, C. et al. 2002, Whiles, M. R. and Charlton, R. E. 2006). However, the bulk of 11

evidence that links grassland habitat structure and arthropod biodiversity is mostly from data 12

obtained in North American, Australian and European ecosystems, and is usually focused on a very 13

specific taxonomic group or guild of organisms (Joern, A. and Laws, A. N. 2013 and references 14

therein, Morris, M. 2000). Arthropods from natural grasslands of South America are considered 15

highly diverse and abundant, but they remain extremely poorly documented and there is still a lack 16

of knowledge regarding their role in ecosystem processes and functioning. 17

This study uses four major arthropod groups (Coleoptera, Araneae, Hemiptera and 18

Formicidae) as model groups to explore the relationship between plant richness and diversity, 19

habitat structure, and arthropod communities in grasslands from southern Brazil. We use vegetation 20

height and plant functional diversity based on life form traits as descriptors of habitat structure. We 21

search for relationships between habitat structure and (1) plant species richness and diversity and 22

(2) arthropod richness and diversity. We also search for relationships between species and 23

functional diversities in both trophic groups. Finally, we explore the relationship between functional 24

traits of arthropods and plants and estimate the magnitude of co-variation between community-25

weighted mean traits using co-inertia analysis. We expect each group to respond differently to 26

Page 98: abordagem quali-quantitativa e funcional de vegetação campestre ...

99

variation on habitat structure, due to contrasting overall habit and feeding behavior (Perner, J. et al. 1

2005). 2

3

Material and Methods 4

5

Study area 6

7

We carried out vegetation and arthropod sampling at five natural grassland sites in Rio Grande do 8

Sul, Southern Brazil. Details on site locations and description of grassland vegetation can be found 9

in Ferreira et al. (unpublished [Capítulo 1]). Grassland ecosystems in this region, known as 10

Campos, are inserted in a transitional zone between tropical and temperate climates. Mean 11

temperatures range from 9oC in winter to 25oC in summer and mean annual rainfall is ca. 1,440 mm 12

mm (Moreno, J. A. 1961). Although it is assumed that climate in the region has no pronounced dry 13

season (Cfa and Cfb in Koppen’s system), there is a high probability of soil water deficit during the 14

peak of summer in the Pampa biome, especially near the Brazilian border with Uruguay (Leivas, J. 15

F. et al. 2006). Grasslands at all sites have been under grazing for many years, with no known 16

record of land conversion. Prior to the surveys we estimated grazing pressure at each site in animal 17

units (AU; corresponding to 450 kg of live weight) per hectare. Grazing pressure in each site was as 18

follow: site 1 = 1.05, site 2 = 0.9, site 3 = 0.6, site 4 = 0.85 and site 5 = 0.45 AU/ha. Campos 19

grasslands are relict ecosystems which covered larger areas in South America in past colder and 20

dryer climates (Behling, H. and Pillar, V. 2008). Present climatic conditions favor forest expansion 21

over grasslands, and this process is kept at bay by fire and grazing disturbances, which in fact 22

maintain grassland physiognomy, structure and diversity (Overbeck, G. E. et al. 2005, Overbeck, G. 23

E. and Pfadenhauer, J. 2007). The Brazilian Campos are the northernmost part of a larger formation 24

known as the Rio de la Plata Grasslands, which cover ca. 750,000 km2 in Argentina, Uruguay and 25

Page 99: abordagem quali-quantitativa e funcional de vegetação campestre ...

100

Southern Brazil (Soriano, A. et al. 1992) and is known for its species richness, ecological relevance 1

and high conservation potential (Bilenca, D. and Miñarro, F. 2004, Overbeck et al. 2007). 2

3

Plant and arthropod sampling 4

5

At each site, we delimited, in 2010, three permanent paddocks of 0.5 ha (70 x 70 m), which 6

resulted in 15 paddocks across five sites. This sampling layout was designed for an ongoing long-7

term ecological research (PELD Campos Sulinos; CNPq 558282/2009-1). We sampled plants and 8

arthropods within each paddock, and used average values per paddock for most analyses (see 9

below). 10

Plant sampling took place in late 2010 and early 2011. Within each paddock, we sampled 11

plant species in nine 1m2 plots, systematically allocated in a 3x3 grid with 17 m between plots (135 12

plots in total). We sampled all plant species present in each plot, and estimated their cover using a 13

decimal scale (Londo, G. 1976). We categorized species in life form categories (Table 1) according 14

to the classification for subtropical grasslands proposed by Ferreira et al. (unpublished [Capítulo 15

2]). In this classification, plant life forms are based on characteristics such as habit, architecture and 16

strategy of horizontal occupation, and ultimately describe vegetation horizontal (and to some degree 17

vertical) structural heterogeneity. We measured vegetation height in five points per plot as a 18

descriptor of vegetation vertical structure and heterogeneity (675 measurements in total). We also 19

estimated cover of bare soil, litter, rock outcrops and overall vegetation cover in each plot. 20

Arthropod community was sampled in each paddock using a sweep net (50 cm large; 0.1 21

m2). We swept the grassland vegetation along four transections equally distributed inside the 22

paddocks. The organisms were preserved in alcohol 80% and sorted in major taxonomical groups. 23

We described each of the focal study groups (Hemiptera, Araneae, Formicidae and Coleoptera) in 24

morphological functional traits, which were measured under stereomicroscope with ocular 25

micrometer. Also, we assigned each group/individual to generalized feeding habits. However, each 26

Page 100: abordagem quali-quantitativa e funcional de vegetação campestre ...

101

group presented peculiarities concerning taxonomical identification and functional trait description. 1

We provide the full description of functional traits in Table 1. 2

Formicidae were identified in genus and morphospecies. Morphological traits (body, 3

relative leg and relative eye size) were measured in up to three individuals of each morphospecies in 4

each paddock. We always selected the smaller individuals, and used the mean of these 5

measurements as the value for each species in each paddock. We selected these morphological traits 6

based on Bihn, J. H. et al. (2010). As a proxy for ant body size we measured head length, taking 7

into account the strong correlation between both structures (Kaspari, M. and Weiser, M. 1999). 8

Relative leg and eye size represented respectively the ratio of leg length (femur and tibia combined) 9

and eye length to head length. Ant feeding behavior information (generalist, predator or leaf cutter) 10

was obtained for all species based on genus level (Brandão, C. R. F. et al. 2012). 11

Coleoptera was identified in family and morphospecies. Morphological traits (body, relative 12

leg and relative elytron size) were obtained from up to three randomly selected individuals from 13

each morphospecies in each paddock, and then we used the average of these values per species per 14

paddock. The pronotum length was used as a proxy for beetle body size, and for relative leg and 15

elytron size we used respectively the ratio of tibia and elytron length to pronotum length. Beetle 16

feeding habits (herbivore or predator) were assigned according to family affiliation (Marinoni, R. et 17

al. 2001). 18

Spiders were identified solely in families, and all sampled individuals from each paddock 19

had morphological traits (body and relative chelicerae size) measured. The body size was described 20

in terms of the cephalothorax area (cephalothorax length x width), and the relative chelicerae size 21

represented the ratio of chelicerae length to cephalothorax length. Spiders are exclusive predators, 22

and we registered their behavior considering the construction of prey-capture web (web builders or 23

hunters) based on family affiliation (Dias, S. C. et al. 2009). 24

Finally, Hemiptera was classified in family, and also all sampled individuals were measured 25

in morphological traits (body and relative stylet size). Body size was estimated based on complete 26

Page 101: abordagem quali-quantitativa e funcional de vegetação campestre ...

102

individual length (excluding wings), and relative stylet size represented the ratio between stylet 1

length and body length. Based on family affiliation, all sampled hemipterans were considered plant-2

sucking herbivores. 3

We assumed the selected morphological traits to be related, in each arthropod group, to 4

their adaptation to the environment, to the relationships with habitat structure and to resource use. 5

Body size is considerate the most fundamental trait of an animal and is usually correlated with life 6

history and physiological and ecological proprieties (e.g., Cushman, J. H. et al. 1993). Relative leg 7

size in ants and beetles may be related to efficiency in locomotion, resource acquisition and may 8

allow or prevent the use of specific microhabitats (Kaspari, M. and Weiser, M. 1999, Wiescher, P. 9

T. et al. 2012). Relative eye size in ants is an important trait related to foraging (Bihn, J. H. et al. 10

2010). Relative elytra size in beetles has been related to dispersal ability (Barton, P. S. et al. 2011). 11

Relative chelicerae size in spiders may be related to prey size and foraging (Podgaiski, L. R. et al. 12

2013) and finally, stylet size in hemipterans is related to sucking efficiency (Grimaldi, D. 2005). 13

The feeding behavioral guilds, and hunting strategies in case of spiders, includes the organisms 14

using the same class of resources, thus reflecting their direct connection with the habitat. 15

16

Data analysis 17

18

Our analyses aimed to (a) search for relationships between vegetation vertical structure, 19

patterns of plant diversity and richness and the arthropod community; (b) estimate the magnitude 20

and significance of the co-variation between plant and arthropod traits and (c) evaluate the role of 21

habitat structure (based on vertical heterogeneity and plant life forms) over four abundant arthropod 22

groups, as well as possible differences among groups regarding this role. 23

We organized the data in matrices containing community information (matrices W of 24

species or individuals by paddocks) and functional information (matrices B of species or individuals 25

by traits). Arthropod data was summarized in two matrices for each group: WH and BH for 26

Page 102: abordagem quali-quantitativa e funcional de vegetação campestre ...

103

Hemiptera, WA and BA for Araneae, WF and BF for Formicidae and WC and BC for Coleoptera. 1

Coleoptera and Formicidae matrices contained information of morpho-species, whereas Araneae 2

and Hemiptera matrices contained information of individuals. Matrices WH and WA are binary, 3

whereas matrices WC and WA contain abundance information (see Table 1 for information on traits 4

describing each matrix B). Plant data comprised two matrices: WP (paddocks described by plant 5

species average cover values) and BP (plant species described by life form categories; see Table 1). 6

All matrices W were standardized to unit total. 7

Using these matrices we calculated community weighted mean traits (CWM; see Podgaiski, 8

L. R. et al. 2013 for details) for each paddock. We generated five matrices T with CWM trait values 9

(one for each group of arthropod and one for plants) by matrix multiplication T = WB (Pillar, V. D. 10

et al. 2009), where W and B were used according to each taxonomic group. The resulting matrices 11

were TP of plant CWM traits and matrices TC, TF, TA and TH of Coleoptera, Formicidae, Araneae 12

and Hemiptera CWM traits, respectively. In cases when a given group of arthropods was absent 13

from a paddock, the vector corresponding to that paddock was also removed from plant matrices 14

WP and BP in order to obtain matrices T with compatible dimensions for both taxonomic groups (to 15

enable co-inertia analysis; see below). 16

We calculated species diversity using Simpson’s 1-D (Magurran, A. E. and McGill, B. J. 17

2011) and functional diversity using Rao’s quadratic entropy (Botta‐Dukát, Z. 2005) for plants and 18

arthropods (for each order separately; species diversity only for Ants and Coleoptera). Rao’s index 19

of functional diversity contains information on the extent of trait dissimilarity among taxa in a given 20

community weighted by their relative abundance. Since plant traits consisted on life form 21

categories, we interpreted plant functional diversity as a measure of habitat heterogeneity. We 22

searched for associations between variables using pairwise correlation analysis with permutation 23

and linear models. Pairwise correlation between vectors from matrices with CWM trait values was 24

also tested to search for relationships between arthropod and plant traits. Possible associations 25

between plant richness, diversity and functional diversity were tested at three different levels (area, 26

Page 103: abordagem quali-quantitativa e funcional de vegetação campestre ...

104

paddock and plot), whereas all other associations were tested at the paddock level. We also 1

searched for relationships between management intensity (grazing pressure) and vegetation 2

structure and diversity. Finally, we calculated plant diversity using multiple diversity indexes 3

(Simpson’s D, Shannon, Evenness, Brillouin, Menhinick, Margalef, Equitability, Fischer-alpha and 4

Berger-Parker; Magurran, A. E. and McGill, B. J. 2011) and estimated the correlation of each index 5

with vegetation structure parameters and richness of arthropod orders. We only discussed 6

significant correlations (P < 0.05), and only presented P values when they were lower than 0.01. 7

We used co-inertia analysis to further explore the relationship between functional traits of 8

arthropods and plants (Dolédec, S. and Chessel, D. 1994). We tested the significance and estimated 9

the magnitude of co-variation between matrix TP of plant CWM traits and matrices TC, TF, TA and 10

TH of arthropod CWM traits. In the next step we carried out Principal Coordinate Analyses (PCA) 11

with all matrices T and reduced their dimensionality by selecting the principal axes. We maximized 12

the concordance between each pair of matrices by rotation of the multivariate ordination space, 13

which generated new axes (Dray, S. et al. 2003). We used permutation to test the significance of 14

each association. All analyses were performed with the R platform (R Development Core Team 15

2012). 16

17

Results 18

19

Overall, across the five sampling sites, we sampled 376 plant species from 40 families and 20

2,579 arthropod individuals from 11 orders. Hemiptera (1,093 individuals), Araneae (500), 21

Formicidae (188) and Coleoptera (141) represented 74% of the overall arthropod abundance 22

sampled in the study. Differences in grazing pressures among sites resulted in differences of 23

vegetation height and variance in vegation height among sites. These differences are, to a large 24

extent, the reflection of different grazing pressures in each site, which influenced both vertical 25

structure (Figure 3a,b) and plant diversity (Figure 3c,d). However, local factors such as soil 26

Page 104: abordagem quali-quantitativa e funcional de vegetação campestre ...

105

composition and large-scale climatic factors may have also contributed to differences in vegetation 1

structure among and within sites and paddocks. 2

Plant species diversity and functional diversity were lower in paddocks from sites 3 and 5 3

(Figure 2). Plant diversity (1-D) and functional diversity (Rao) were positively correlated at 4

paddock level (r = 0.89, Figure 4b), but not correlated at site and plot levels. Similarly, plant species 5

richness and functional diversity were positively correlated at paddock level (r = 0.81, p <0.01), but 6

not correlated at site and plot levels. All plant diversity indexes were negatively correlated with 7

average vegetation height and positively correlated with height variance and richness of arthropod 8

orders (inverse relationships for Simpson’s D and Berger-Parker’s index). 9

Average vegetation height and plant species diversity were negatively correlated (r = -0.78, 10

P < 0.01), as well as average vegetation height and plant functional diversity (r = -0.72, Figure 4a). 11

Vegetation height variance and plant diversity were positively correlated (r = 0.78). Average 12

percentage of bare soil and plant diversity were positively correlated (r = 0.88). Richness of 13

arthropod orders was positively correlated to plant diversity (r = 0.63) and plant functional diversity 14

(r = 0.73; Figure 5a). Plant functional diversity was also positively correlated to Coleoptera 15

functional diversity (Figure 5b). 16

Co-inertia analyses showed a significant association between matrices T of plants and 17

arthropods CWM trait values for all arthropod orders (Table 2). The association was stronger in 18

Coleoptera and Araneae (RV = 0.683 and 0.438, respectively), and weaker in Hemiptera and 19

Formicidae (RV = 0.332). The first two axes of the co-inertia biplot for each group of arthropods 20

represented the differences in vegetation structure across the sampling sites (Figures 6a,d and 7a,d). 21

Paddocks with overall higher and more homogeneous vertical vegetation structure (Figure 1) and 22

lower values of plant diversity and functional diversity (Figure 2) were segregated from paddocks 23

with lower and more heterogeneous vegetation, with higher plant diversity. This pattern was more 24

evident for Coleoptera and Araneae (Figure 6) and less evident for Hemitera and Formicidae 25

(Figure 7). It is likely that these differences are the reflection of different relationships between 26

Page 105: abordagem quali-quantitativa e funcional de vegetação campestre ...

106

habitat structure and arthropod community in each group. Overall, paddocks from the same sites did 1

not group together in the biplots, suggesting that vegetation structure is more important as a 2

defining factor for arthropod community structure than spatial proximity alone. 3

Pairwise tests between vectors representing CWM traits for plants and arthropods (from 4

matrices T) resulted in several significant associations. For spiders, ‘stoloniferous plants’ showed 5

association with hunters (r = 0.71) and web-builders (r = -0.71); ‘subshrubs’, ‘forbs’ and 6

‘decumbent plants’ showed positive or negative association with relative size of chelicera (r = -0.54, 7

0.60 and 0.53, respectively; Figure 6b,c). For coleopterans, ‘tussocks’ was negatively associated 8

with body size (r = -0.66), predators (r = 0.71) and herbivores (r = -0.71); ‘forbs’ was associated 9

with body size (r = 0.69) and ‘subshrubs’ was associated with predators (r = 0.72), elytron size (r = 10

0.48) and leg size (r = 0.74; Figure 6e,f). Regarding ants, ‘therophytes’ was associated with leaf-11

cutters (r = 0.58), and ‘shrubs’ with body size (r = -0.64 , Figure 7b,c). For hemipterans, 12

‘stoloniferous plants’ showed negative association with body size (r = -0.78) and positive 13

association with mouthparts (r = 0.65); ‘subshrubs’ showed negative association with mouthparts (r 14

= -0.54) and positive association with body size (r = 0.73; Figure 7e,f). Finally, spider functional 15

diversity and ‘erect rosettes’ were positively correlated (r = 0.66), as well as hemiptera and 16

coleoptera functional diversities and ‘shrubs’ (r = 0.79 and 0.67, respectively). 17

18

Discussion 19

20

Our objective was to explore the relationships between vegetation structure and (1) plant 21

diversity and (2) arthropod diversity in grasslands from Southern Brazil. Although our results 22

represent a snapshot of these relationships, which may vary in time, we did find some clear patterns 23

linking vegetation structure, plant diversity and arthropods. Moreover, our results provide further 24

evidence that grassland vegetation structure is strongly associated with management, as differences 25

in grazing pressures between our sampling sites are associated with variations in vertical vegetation 26

Page 106: abordagem quali-quantitativa e funcional de vegetação campestre ...

107

structure and plant diversity (Figure 2). The role of disturbance (grazing and/or fire) on grassland 1

vegetation heterogeneity is widely accepted (Bond, W. et al. 2005, Bond, W. J. and Keeley, J. E. 2

2005, Knapp, A. K. et al. 1998), but supporting evidence for this relationship, especially regarding 3

the role of grazing, from subtropical and temperate grasslands is still scarce. Light grazing pressures 4

promote the predominance of tussock grasses and the development of shrubs, which outcompete 5

most prostrate life forms and ultimately reduce plant diversity. Our findings corroborate that 6

relationship: vegetation height and height variance decrease linearly with decreasing grazing 7

pressures (Figure 2a,b), and this pattern was recurrent for plant diversity (Figures 2, 3 and 4a). In a 8

local scale, overgrazing may produce the same effect by creating uniform ‘prostrate plant 9

communities’, although this was not the case in any of our sampling sites. 10

We used plant life forms as traits (Table 1) to calculate plant functional diversity (FD), as a 11

proxy of habitat heterogeneity. The FD index we used estimates trait dissimilarity among taxa in the 12

communities (each paddock) weighted by their relative abundances (Botta‐Dukát, Z. 2005). 13

Paddocks with low FD are largely dominated by few plant life forms, whereas in paddocks with 14

high FD plant cover (our proxy for abundance) is more evenly distributed among different life 15

forms, which promotes higher habitat heterogeneity. 16

Grassland arthropods are strongly influenced by habitat structure, which in turn is defined 17

by plant species (Joern, A. and Laws, A. N. 2013). Evidence linking arthropod abundances to plant 18

species richness is plentiful, but results encompass both positive and negative relationships 19

(Koricheva, J. et al. 2000), and functional aspects of both taxa are mostly ignored. Moreover, 20

studies that pursue plant/arthropod relationships often focus on small groups or single species. Our 21

results indicate that areas with more heterogeneous habitats (i.e., with higher plant richness and FD) 22

encompass increased richness of arthropod orders (Figure 5A). Structurally complex and diverse 23

grassland areas provide an increased variety of niches and resources to be exploited in comparison 24

with more homogeneous areas (Bazzaz, F. 1975), which should enable the existence of an arthropod 25

community with more diverse habitat and resource requirements. Further, FD of spiders and beetles 26

Page 107: abordagem quali-quantitativa e funcional de vegetação campestre ...

108

were also clearly positively related to plant FD. Podgaiski, L. R. et al. (2013) found similar results 1

for spiders in grasslands from south Brazil in an experiment on effects of fire: more functionally 2

diverse plant communities (induced by fire disturbance) sheltered more functionally diverse spider 3

communities. 4

The functional traits of the different arthropod orders and trophic groups were affected by 5

different components of grassland plant diversity and showed some interesting but yet unclear 6

correlations with plant life form traits. For example, body size in each group was affected by 7

different plant traits: body size of Hemiptera was influenced by subshrubs and stoloniferous plants; 8

in Coleoptera it was affected by tussock grasses and forbs, while in ants it was more influenced by 9

the proportion of shrubs in vegetation cover. Considering the size-grain hypothesis (Kaspari, M. and 10

Weiser, M. 1999), morphological features of walking organisms (such as body size) indicate a 11

trade-off in the manner they move over and through the environment and use the habitat. Smaller 12

organisms experience a more rugose world with the possibility of enhanced penetration on 13

microhabitats and interstices that larger organisms are not suited to enter (Sarty, M. et al. 2006). In 14

our results, this could be noticed in the relationship between Coleoptera and habitat structure: 15

coleopteran body size decreased following an increasing trend in the proportions of tussock grasses. 16

The predominance of tussock grasses shapes habitats with high density of leaves, which would be 17

best suited for smaller coleopterans. 18

Vegetation structure is particularly important for generalist predators such as spiders 19

(Podgaiski, L. R. et al. 2013, Sunderland, K. and Samu, F. 2000). Non-web-building spiders 20

(cursorials) have been shown to benefit in habitats with higher proportions of prostrate plant 21

species, i.e. habitats that do not provide the necessary architecture to allow web building. Also, 22

hunting spiders with larger chelicerae could benefit from such clear environments (Podgaiski, L. R. 23

et al. 2013). 24

Our results showed that leaf-cutting ants were associated with higher proportion of annual 25

plants. It has been shown that leafcutters may select pioneer leaves because of their low level of 26

Page 108: abordagem quali-quantitativa e funcional de vegetação campestre ...

109

chemical defenses and high nutrient content (Farji‐Brener, A. G. 2001), and our results confirm this 1

‘palatable forage hypothesis’, as annual plant species usually are poor in fiber and show high 2

nutritious value. 3

4

Conclusions 5

6

Our results provide further evidence supporting the close relationship between disturbance 7

(grazing) and grassland vegetation structure and diversity patterns. Although this relationship is 8

well established, empirical evidence from subtropical grasslands is still scarce. Grazing promotes 9

habitat heterogeneity, which in turn influences diversity patterns of arthropod. Although our results 10

regarding the relationships between arthropods and vegetation structure are on the coarse level of 11

four major orders, they may guide future directions for research on links between grassland 12

vegetation and specific arthropod groups in subtropical grasslands, choosing for example animal 13

traits that are more responsive to habitat structure in each group. 14

15

Acknowledgements 16

17

We thank all farm owners that kindly allowed this work to be carried out in their properties, and all 18

local environmental authorities that allowed research in protected areas and helped us during 19

fieldwork. The first author thanks CAPES for a scholarship. I. I. Boldrini thanks CNPq for a 20

research productivity grant.21

Page 109: abordagem quali-quantitativa e funcional de vegetação campestre ...

110

Table 1. Description of arthropod and plant traits used in the study.

Group Trait category Trait Acronym Category Description

Hemiptera Morphological Relative stylet size Mouthp Quantitative Stylet (mm) / body lenght (mm)

Morphological Body size Body Quantitative Body lenght (mm)

Araneae Feeding behavior Hunter Hunt Binary

Feeding behavior Web builder Web Binary

Morphological Body size Body Quantitative Area of cephalothorax (mm2)

Morphological Relative chelicera lenght Chel Quantitative Chelicerae length (mm) / cephalothorax length (mm)

Formicidae Feeding behavior Generalist Gen Binary

Feeding behavior Leaf cutter Cutter Binary

Feeding behavior Predator Pred Binary

Morphological Body size Body Quantitative Head lenght (mm)

Morphological Relative leg size Leg Quantitative Leg length (mm) / head length (mm)

Morphological Relative eye size Eye Quantitative Eye length (mm) / head length (mm)

Coleoptera Feeding behavior Herbivore Herb Binary

Feeding behavior Predator Pred Binary

Morphological Body size Body Quantitative Pronoto lenght (mm)

Morphological Relative leg size Leg Quantitative Tibia lenght (mm) / pronoto lenght (mm)

Morphological Relative elytron size Ely Quantitative Elytron length (mm) / pronoto length (mm)

Plants Life form Therophytes Th Binary

Life form Bulbous geophytes Bg Binary

Life form Rhizomatous geophytes Rg Binary

Life form Prostate rosette evergreens Pr Binary

Life form Decumbent evergreens De Binary

Life form Rhizomatous evergreens Rh Binary

Life form Stolonoiferous evergreens St Binary

Life form Solitary evergreen tussocks Te Binary

Life form Connected evergreen tussocks Ct Binary

Life form Evergreen forbs Ef Binary

Life form Erect rosette evergreens Er Binary

Life form Evergreen subshrubs Ss Binary

Life form Evergreen shrubs Sh Binary

Page 110: abordagem quali-quantitativa e funcional de vegetação campestre ...

111

Table 2. Summarized results of co-inertia analysis using community-weighted mean traits of plants

and each arthropod group.

Hemiptera Araneae Formicidae Coleoptera

RV 0.332 0.438 0.332 0.683

P-value 0.034 0.017 0.028 0.001

Axis 1 (%) 97.5 79.1 64.9 93.8

Axis 2 (%) 2.1 17.2 19.6 5.1

Cumulative (%) 99.6 96.3 84.5 98.9

Figure 1. Vertical structure of grassland vegetation across five sampling sites (1 - 5) and 15

paddocks (a, b and c), nine sampling units per paddock.

Page 111: abordagem quali-quantitativa e funcional de vegetação campestre ...

112

Figure 2. Plant species diversity (1-D) and functional diversity (Rao’s quadratic entropy) across

five sampling sites (1 - 5) and 15 paddocks (a, b and c).

Figure 3. Relationships between grazing pressure and (A) vegetation height, (B) vegetation height

variance, (C) plant species diversity and (D) plant functional diversity. UA = animal units (450 kg

of live weight).

0.50

0.60

0.70

0.80

0.90

1.00

a b c a b c a b c a b c a b c

1 2 3 4 5

Pla

nt

div

ers

ity

Sites/paddocks

1-D

RAO

Page 112: abordagem quali-quantitativa e funcional de vegetação campestre ...

113

Figure 4. Relationships between vegetation vertical structure and plant functional diversity (A) and

plant species and functional diversities (B). Labels: 15 paddocks (a, b and c) from five sampling

sites (1 - 5).

1a 1b 1c

2a 2b

2c

3a

3b 3c

4a

4b 4c 5a 5b

5c

R² = 0.84822 0.50

0.60

0.70

0.80

0.90

1.00

0.50 0.60 0.70 0.80 0.90 1.00

Pla

nt

sp

ec

ies d

ive

rsit

y (

1-D

)

Plant functional diversity (Rao)

1a 1b

1c

2a 2b 2c

3a 3b

3c

4a

4b 4c

5a

5b 5c

R² = 0.39532 P = 0.012

P < 0.001

0

5

10

15

20

25

30

35

40

0.5 0.6 0.7 0.8 0.9 1

Av

era

ge v

eg

eta

tio

n h

eig

ht

(cm

)

Plant functional diversity (Rao)

A

B

Page 113: abordagem quali-quantitativa e funcional de vegetação campestre ...

114

Figure 5. Relationships between plant functional diversity and: (A) richness of arthropod orders;

(B) coleopteran functional diversity and (C) spider functional diversity. Labels: 15 paddocks (a, b

and c) from five sampling sites (1 - 5).

Page 114: abordagem quali-quantitativa e funcional de vegetação campestre ...

115

Figure 6. Results of co-inertia analysis using community-weighted mean traits of plants and spiders

(A-C) and coleopterans (D-F). A and D: ordination of paddocks (a, b and c) from five sampling

sites (1 - 5) based on plant (full circles) and arthropod (empty circles) traits. PCA of plant (B and E)

and arthropod (C and F) traits. Highlighted traits indicate significant pairwise correlation (same

symbol) between plant and arthropod traits within the same group. Legend for traits: see Table 1.

Page 115: abordagem quali-quantitativa e funcional de vegetação campestre ...

116

Figure 7. Results of co-inertia analysis using community-weighted mean traits of plants and ants

(A-C) and trips (D-F). A and D: ordination of paddocks (a, b and c) from five sampling sites (1 - 5)

based on plant (full circles) and arthropod (empty circles) traits. PCA of plant (B and E) and

arthropod (C and F) traits. Highlighted traits indicate significant pairwise correlation (same symbol)

between plant and arthropod traits within the same group. Legend for traits: see Table 1.

Page 116: abordagem quali-quantitativa e funcional de vegetação campestre ...

117

References

Barton, P. S. et al. 2011. Morphological traits as predictors of diet and microhabitat use in a

diverse beetle assemblage. — Biological journal of the Linnean Society 102: 301-310.

Bazzaz, F. 1975. Plant species diversity in old-field successional ecosystems in southern Illinois.

— Ecology 485-488.

Behling, H. and Pillar, V. 2008. Vegetation and fire dynamics in southern Brazil during the late

Quaternary and their implication for conservation and management of modern grassland

ecosystems. — In: Schröder, H. (ed), Grasslands: Ecology, Management and Restoration. Nova

Science Publishers, pp. 99-108.

Behling, H. et al. 2004. Late Quaternary Araucaria forest, grassland (Campos), fire and climate

dynamics, studied by high-resolution pollen, charcoal and multivariate analysis of the Cambará do

Sul core in southern Brazil. — Palaeogeography, Palaeoclimatology, Palaeoecology 203: 277-297.

Belovsky, G. and Slade, J. 2000. Insect herbivory accelerates nutrient cycling and increases plant

production. — Proceedings of the National Academy of Sciences 97: 14412-14417.

Bihn, J. H. et al. 2010. Loss of functional diversity of ant assemblages in secondary tropical

forests. — Ecology 91: 782-792.

Bilenca, D. and Miñarro, F. 2004. Identificación de áreas valiosas de pastizal en las pampas y

campos de Argentina, Uruguay y sur de Brasil. — Fundación Vida Silvestre

Bond, W. et al. 2005. The global distribution of ecosystems in a world without fire. — New

Phytologist 165: 525-538.

Bond, W. J. and Keeley, J. E. 2005. Fire as a global ‘herbivore’: the ecology and evolution of

flammable ecosystems. — Trends in Ecology & Evolution 20: 387-394.

Botta‐Dukát, Z. 2005. Rao's quadratic entropy as a measure of functional diversity based on

multiple traits. — Journal of Vegetation Science 16: 533-540.

Page 117: abordagem quali-quantitativa e funcional de vegetação campestre ...

118

Brandão, C. R. F. et al. 2012. Neotropical Ants (Hymenoptera) Functional Groups. — In:

Panizzi, A. R. and Parra, J. R. (eds), Insect Bioecology and Nutrition for Integrated Pest

Management. CRC Press, pp. 213–236.

Cushman, J. H. et al. 1993. Latitudinal patterns in European ant assemblages: variation in

species richness and body size. — Oecologia 95: 30-37.

Dias, S. C. et al. 2009. Refining the establishment of guilds in Neotropical spiders (Arachnida:

Araneae). — Journal of Natural History 44: 219-239.

Diaz, S. et al. 2007. Plant trait responses to grazing–a global synthesis. — Global Change

Biology 13: 313-341.

Dolédec, S. and Chessel, D. 1994. Co‐inertia analysis: an alternative method for studying

species–environment relationships. — Freshwater biology 31: 277-294.

Dray, S. et al. 2003. Co-inertia analysis and the linking of ecological data tables. — Ecology 84:

3078-3089.

Farji‐Brener, A. G. 2001. Why are leaf‐cutting ants more common in early secondary forests

than in old‐growth tropical forests? An evaluation of the palatable forage hypothesis. — Oikos 92:

169-177.

Fay, P. A. 2003. Insect diversity in two burned and grazed grasslands. — Environmental

Entomology 32: 1099-1104.

Fuhlendorf, S. D. and Engle, D. M. 2001. Restoring Heterogeneity on Rangelands: Ecosystem

Management Based on Evolutionary Grazing Patterns: We propose a paradigm that enhances

heterogeneity instead of homogeneity to promote biological diversity and wildlife habitat on

rangelands grazed by livestock. — BioScience 51: 625-632.

Grimaldi, D. 2005. Evolution of the Insects. — Cambridge University Press.

Grime, J. P. 2006. Trait convergence and trait divergence in herbaceous plant communities:

mechanisms and consequences. — Journal of Vegetation Science 17: 255-260.

Page 118: abordagem quali-quantitativa e funcional de vegetação campestre ...

119

Joern, A. and Laws, A. N. 2013. Ecological mechanisms underlying arthropod species diversity

in grasslands. — Annual review of entomology 58: 19-36.

Kaspari, M. and Weiser, M. 1999. The size–grain hypothesis and interspecific scaling in ants. —

Functional Ecology 13: 530-538.

Knapp, A. K. et al. 1998. Grassland dynamics: long-term ecological research in tallgrass prairie.

— Oxford University Press New York.

Koricheva, J. et al. 2000. Numerical responses of different trophic groups of invertebrates to

manipulations of plant diversity in grasslands. — Oecologia 125: 271-282.

Kruess, A. and Tscharntke, T. 2002. Contrasting responses of plant and insect diversity to

variation in grazing intensity. — Biological conservation 106: 293-302.

Leivas, J. F. et al. 2006. Risco de deficiência hídrica decendial na metade sul do Estado do Rio

Grande do Sul. — Revista Brasileira de Engenharia Agrícola e Ambiental 10: 397-407.

Londo, G. 1976. The decimal scale for relevés of permanent quadrats. — Vegetatio 33: 61-64.

Loreau, M. et al. 2002. Biodiversity and ecosystem functioning. — Oxford University Press

Oxford.

MacArthur, R. H. 1967. The theory of island biogeography. — Princeton University Press.

Magurran, A. E. and McGill, B. J. 2011. Biological diversity: frontiers in measurement and

assessment. — Oxford University Press Oxford.

Marinoni, R. et al. 2001. Hábitos alimentares em Coleoptera (Insecta). — Ribeirão Preto, Holos,

63p.[Links]

McCoy, E. and Bell, S. 1991. Habitat structure: the evolution and diversification of a complex

topic. — Habitat structure 3-27.

Meyer, C. et al. 2002. Life history, secondary production, and ecosystem significance of acridid

grasshoppers in annually burned and unburned tallgrass prairie. — American Entomologist 48: 52-

61.

Page 119: abordagem quali-quantitativa e funcional de vegetação campestre ...

120

Milchunas, D. et al. 1988. A generalized model of the effects of grazing by large herbivores on

grassland community structure. — American Naturalist 87-106.

Moreno, J. A. 1961. Clima do Rio Grande do Sul. — Secretaria da Agricultura.

Morris, M. 2000. The effects of structure and its dynamics on the ecology and conservation of

arthropods in British grasslands. — Biological Conservation 95: 129-142.

Overbeck, G. E. et al. 2005. Fine‐scale post‐fire dynamics in southern Brazilian subtropical

grassland. — Journal of Vegetation Science 16: 655-664.

Overbeck, G. E. and Pfadenhauer, J. 2007. Adaptive strategies in burned subtropical grassland in

southern Brazil. — Flora-Morphology, Distribution, Functional Ecology of Plants 202: 27-49.

Perner, J. et al. 2005. Effects of plant diversity, plant productivity and habitat parameters on

arthropod abundance in montane European grasslands. — Ecography 28: 429-442.

Pillar, V. D. et al. 2009. Discriminating trait‐convergence and trait‐divergence assembly patterns

in ecological community gradients. — Journal of Vegetation Science 20: 334-348.

Podgaiski, L. R. et al. 2013. Spider Trait Assembly Patterns and Resilience under Fire-Induced

Vegetation Change in South Brazilian Grasslands. — PloS one 8: e60207.

R Development Core Team 2012. R: A Language and Environment for Statistical Computing. R

Foundation for Statistical Computing.

Sarty, M. et al. 2006. Habitat complexity facilitates coexistence in a tropical ant community. —

Oecologia 149: 465-473.

Simpson, E. H. 1949. Measurement of diversity. — Nature 163: 688.

Soriano, A. et al. 1992. Río de la Plata Grasslands. In ‘Ecosystems of the world 8A. Natural

grasslands. Introduction and Western Hemisphere’.(Ed. RT Coupland) pp. 367–407. Elsevier:

Amsterdam.

Sunderland, K. and Samu, F. 2000. Effects of agricultural diversification on the abundance,

distribution, and pest control potential of spiders: a review. — Entomologia Experimentalis et

Applicata 95: 1-13.

Page 120: abordagem quali-quantitativa e funcional de vegetação campestre ...

121

Tews, J. et al. 2004. Animal species diversity driven by habitat heterogeneity/diversity: the

importance of keystone structures. — Journal of Biogeography 31: 79-92.

Tscharntke, T. and Greiler, H.-J. 1995. Insect communities, grasses, and grasslands. — Annual

Review of Entomology 40: 535-558.

Vandewalle, M. et al. 2010. Functional traits as indicators of biodiversity response to land use

changes across ecosystems and organisms. — Biodiversity and Conservation 19: 2921-2947.

Wardle, D. A. et al. 2005. Trickle‐down effects of aboveground trophic cascades on the soil food

web. — Oikos 111: 348-358.

Whiles, M. R. and Charlton, R. E. 2006. The ecological significance of tallgrass prairie

arthropods. — Annu. Rev. Entomol. 51: 387-412.

Wiescher, P. T. et al. 2012. Assembling an ant community: species functional traits reflect

environmental filtering. — Oecologia 169: 1063-1074.

Page 121: abordagem quali-quantitativa e funcional de vegetação campestre ...

122

Page 122: abordagem quali-quantitativa e funcional de vegetação campestre ...

123

Title: Long-term ecological research in subtropical grasslands: results from a four-year monitoring 1

of different management practices in Southern Brazil 2

3

Pedro M.A. Ferreira, Bianca O. Andrade, Luciana R. Podgaiski, Gerhard E. Overbeck, Ilsi I. 4

Boldrini 5

6

Abstract 7

8

Questions: How do subtropical grassland communities respond to different intensities of grazing 9

disturbance through time? Does this response affects litter decomposition? 10

11

Location: Subtropical grasslands, Southern Brazil (27o15’S - 31o54’S; 50o15’W - 56o15’W; 150 – 12

850 masl). 13

14

Method: We selected six sites with natural grassland vegetation under grazing. In each site we 15

delimited three paddocks to which we randomly assigned one of three different managements with 16

grazing animals: conventional, conservative or exclusion. The conventional treatment was our 17

control, in which we maintained the grazing pressure used at each site. The conservative was a 18

simulation of rotational grazing and the exclusion represented the cessation of disturbance. We 19

repeatedly sampled plant communities using permanent plots in south hemisphere spring during 20

four years. We searched for differences between treatments along time in species richness, 21

diversity, functional diversity and functional redundancy. We used an alternative classification of 22

plant life forms as functional traits, and also evaluated shifts in dominance of life forms between 23

treatments. Finally, we estimated litter decomposition in two different experiments using litter-bags. 24

25

Page 123: abordagem quali-quantitativa e funcional de vegetação campestre ...

124

Results: Species richness decreased in exclusion paddocks after the second year of sampling. 1

Management cessation promoted a short-term increase in species diversity. After the second year, 2

species diversity dropped steadily in the exclusion, whereas it remained unchanged under the 3

conservative management. Functional diversity also dropped after the second year of exclusion, but 4

increased in the following years in the conservative treatment. Functional redundancy varied little in 5

the control and conservative, whereas it increased linearly in the exclusion along the four years of 6

sampling. Relations of dominance of life forms shifted in the exclusion and conservative treatments, 7

with sharp decrease in cover of prostrate plants and increase in other life forms. Mean vegetation 8

height and dead biomass increased more than two times in the exclusion. Many species were 9

outcompeted due to shading. Decomposition experiments showed differences between treatments 10

only after the third year of sampling, with higher decomposition rates in conventional paddocks and 11

in Pampa sites. We concluded that these differences were mostly due to differential 12

photodegradation promoted by differences in vegetation structure. 13

14

Conclusion: Interruption of grazing disturbance caused drastic changes in community parameters 15

by shifting relationships of dominance, removing grazing-tolerant species and benefitting groups of 16

species that were controlled by grazing. Structural differences arising from management exclusion 17

also influenced litter decomposition. Our findings suggest that subtropical grasslands from South 18

America may be in an intermediate position in a gradient of resistance/resilience to grazing in 19

comparison with other systems such as shortgrass steppes and tallgrass prairies. 20

21

Keywords: diversity, grazing, Pastizales del Rio de la Plata, plant life forms, litter decomposition, 22

ecosystem processes 23

Page 124: abordagem quali-quantitativa e funcional de vegetação campestre ...

125

Introduction 1

Grasslands, at least when under productive climatic conditions, are disturbance-prone 2

ecosystems strongly shaped by fire and grazing regimes (Milchunas et al. 1988; Knapp et al. 1998; 3

Bond & Keeley 2005). Disturbance can be defined as ‘any event in time that disrupts ecosystem, 4

community, or population structure and changes resource pools, substrate availability, or the 5

physical environment’ (Pickett 1985) or, more simply, as ‘any event partially or totally destroying 6

plant biomass’ (Grime 1979). Either way, disturbance plays a key role on grassland species 7

composition, diversity patterns on multiple scales and ecosystem functioning (e.g., Milchunas et al. 8

1988; Frank & McNaughton 1993; Frank & Evans 1997; Knapp et al. 1998; Frank et al. 2000; Bond 9

et al. 2005; Grime 2006; Diaz et al. 2007). 10

Grazing animals can be very selective as to what they forage (Senft et al. 1987). In 11

productive grasslands, where high-quality palatable plants can be found, herbivores preferably 12

graze these plants, avoiding patches dominated by tall, less palatable taxa (Adler et al. 2001; 13

McIntyre & Tongway 2005; McIvor et al. 2005). This selection promotes heterogeneity by creating 14

a mosaic of patches under different grazing pressures in the landscape with the selection of plants 15

that share traits compatible with each local disturbance situation (Grime 2006; Diaz et al. 2007). 16

However, this selectiveness also depends on the amount of forage available per animal unit in a 17

given paddock: the lower the forage availability, the less selective foraging will be , and the 18

aforementioned patchiness may be converted in structural homogeneity (Senft et al. 1987; 19

Coughenour 1991). The opposite also holds true: high forage availability may lead to higher 20

selectiveness and to increasing dominance of plants characteristic of ungrazed or lightly grazed 21

patches (Hobbs & Swift 1988). 22

To evaluate disturbance-driven grassland heterogeneity, or to compare shifts in grassland 23

communities under different grazing intensities, plant species alone may not be the best working 24

units. To answer such questions, the usefulness of a functional approach is well-established (Diaz et 25

al. 2007). Based on recurrent findings of correlated plant traits, it has been suggested that reduced 26

Page 125: abordagem quali-quantitativa e funcional de vegetação campestre ...

126

sets of traits such as life/growth forms may be good descriptors and predictors of ecosystem 1

functioning under disturbance (McIntyre et al. 1995) or climate change (Chapin 1993). Although 2

plant life forms sensu Raunkiaer (1934) may not be the ideal descriptors in all grassland ecosystems 3

(Ferreira et al. unpublished [Capítulo 2]), alternative case-specific life/growth form classifications 4

have been successfully used to answer different ecological questions in different places around the 5

globe (e.g., Arnold 1955; Grime 1973; Sala 1988; Grime et al. 1997; Hadar et al. 1999). 6

Grazing by large herbivores also affects grassland ecosystem processes such as biomass 7

decomposition. By modifying community habitat structure, plant species and functional 8

composition, herbivores influence biological, physical and chemical properties of the soil 9

environment, which ultimately enhances or reduces plant litter quality and decomposition rates 10

(Bardgett & Wardle 2003; Güsewell et al. 2005; Semmartin et al. 2008). Habitat heterogeneity 11

promoted by grazing can influence biological activity at the soil level by altering microclimatic 12

features (Throop & Archer 2007; Araujo et al. 2012) or facilitating litter photodegradation (Verhoef 13

et al. 2000; Pancotto et al. 2005). Large grazing animals also contribute with decomposable 14

resources and nutrients by depositing urine and faeces, which may lead to increases in populations 15

of decomposition microorganisms and also influence overall decomposition rates (Ruess & 16

McNaughton 1987; Seagle et al. 1992; Bakker et al. 2004). Thus, it is expected that litter 17

decomposition would be accelerated in grazed areas/patches in comparison with ungrazed or lightly 18

grazed areas (e.g., Augustine & McNaughton 1998). However, this assumption did not hold true in 19

South American temperate grasslands: Vaieretti et al. (2013) found no differences in decomposition 20

rates comparing grazed and ungrazed patches in Argentina. Also working in Argentina, Carrera et 21

al. (2008) found that leaf litter decomposed fast at the grazed sites due to changes in canopy 22

structure induced by grazing disturbance. The relationship between grazing intensity and its 23

implications on decomposition seems to remain unclear, especially considering South American 24

subtropical grasslands, were few works addressing the question have been carried out, although 25

there is evidence that fire affects litter decomposition in these ecosystems (Podgaiski et al. 2014). 26

Page 126: abordagem quali-quantitativa e funcional de vegetação campestre ...

127

Although the general importance of grazing for origin and maintenance of grasslands is 1

well established, the impact of this disturbance on ecosystem structure and processes may vary with 2

geographical location. The classic ‘generalized grazing model’ proposed by Milchunas et al. (1988) 3

ascribe these differences in sensibility to grazing intensity (using plant diversity as indicator) to two 4

principal variables: evolutionary history of grazing and moisture level of a given grassland 5

ecosystem. Cingolani et al. (2005) provide further discussion of Milchuna’s model and suggests 6

modifications that enhance its applicability. In another perspective, differences in ‘sensibility’ to 7

disturbance may be related to two different ecological properties: resistance and resilience (Harrison 8

1979). Empirical evidence indicate that shortgrass steppes from Southern South America are highly 9

resistant to grazing (Milchunas & Lauenroth 1993; Milchunas et al. 1998), whereas North American 10

tallgrass prairies are less resistant but more resilient to grazing (Coffin et al. 1996; Baer et al. 2000). 11

The relationships between grazing, vegetation dynamics and ecosystem processes are still 12

poorly studied in ecosystems from Southern Brazil, especially considering long-term monitoring. 13

Grasslands in this region, locally known as ‘Campos’, are relict ecosystems from drier and cooler 14

periods that are stabilized until today by the action of herbivores and fire (Behling & Pillar 2008). 15

There is evidence of the presence of large grazing herbivores in South American grasslands since 16

the early Miocene (MacFadden 1997, 2005). After their extinction, grazing by domestic herbivores 17

has become widespread since the seventeenth century (Porto 1954), and today cattle breeding is one 18

of the most important economic activities in the region (Pillar 2009). However, the degree of 19

resistance/resilience of these systems to grazing remains unclear, as do the consequences of 20

interrupting the disturbance regime. To answer such questions, long-term ecological research 21

(LTER) is essential. Although LTER has greatly improved our understanding of ecosystem 22

dynamics (reviews in Rees et al. 2001; Turner et al. 2003), very little of this evidence comes from 23

the Southern Hemisphere, and almost none from grasslands in Southern South America (but see 24

Boldrini & Eggers 1996). 25

Page 127: abordagem quali-quantitativa e funcional de vegetação campestre ...

128

In this paper we present the results of the first four years of an experiment in LTER sites 1

established in 2010 in the South Brazilian grasslands. We examined shifts in grassland vegetation 2

diversity patterns, relative dominance of plant life forms and biomass decomposition rates under 3

different management intensities with grazing animals. We hypothesized that exclusion of 4

management will lead to a decrease in species richness, diversity and functional diversity (as seen 5

after fire disturbance in similar ecosystems; Overbeck et al. 2005). Accordingly, we expect that 6

different grazing intensities will lead to shifts in dominance of plant life forms. Also, we expect that 7

‘lighter’ grazing pressures under rotational grazing will promote higher heterogeneity and increased 8

species richness, diversity and functional diversity. 9

10

Material and Methods 11

12

Study sites and sampling design 13

14

Our study area comprised six sites in Southern Brazil. Since grassland ecosystems in the 15

region are present in two different biomes, Pampa and Atlantic Forest, we selected three sites in 16

each biome. Grasslands in the Pampa biome cover large continuous areas, and forests are mostly 17

restricted to rivers. In the Atlantic Forest biome grasslands and forests shape mosaics in the 18

landscape (Boldrini 1997; Boldrini et al. 2009). Pampa sites were Aceguá (31o38’55”S, 19

54o09’26”W), Alegrete (30o04’08”S, 55o59’27”W) and Lavras do Sul (30o41’55”S, 53o58’11”W). 20

Atlantic Forest sites were Aparados da Serra National Park (29o08’10”S, 50o09’21”W), Aratinga 21

Ecological Station (29o23’31”S, 50o14’30”W) and Tainhas State Park (29o05’40”S, 50o22’03”W). 22

Grasslands at all sites are under cattle grazing probably since the introduction of domestic cattle in 23

the 17th century in the Pampa and 18th century in the Atlantic Forest. 24

The experiment consisted in a randomized block design. At each site, we delimited three 25

paddocks of 0.5 ha, and to each paddock we randomly assigned one of three treatments that 26

Page 128: abordagem quali-quantitativa e funcional de vegetação campestre ...

129

represented different management of grazing animals: conventional, conservative or exclusion. The 1

conventional treatment was the control, in which we maintained the grazing pressure (animal units 2

per hectare) currently used at a given site. The exclusion treatment consisted on fenced paddocks 3

that completely precluded the entry of grazing animals. The conservative treatment consisted on 4

fenced paddocks with gates that allowed for controlled grazing during certain periods of time, 5

simulating rotational grazing. In the interval between grazing events the conservative paddocks 6

remained inaccessible to grazing animals. In the conservative management, the criterion used to 7

determine the interval of cattle access to each paddock was the accumulated thermal sum of 700 8

degrees-day per site. The duration and number of animals used in each grazing event was calculated 9

to obtain a post-grazing aboveground residual of ca. 1,200 kg of dry biomass per hectare. These 10

procedures aim to maintain the contribution of resource-conserving grasses in the grassland 11

community (Quadros et al. 2006), to promote habitat heterogeneity and less accumulation of dead 12

biomass and ultimately to enhance ecosystem resilience (Soussana 2009). 13

We sampled the grassland vegetation in each site for the first time during south hemisphere 14

spring/summer 2010, after which we carried out the construction of fences to start the different 15

treatments. Cattle access into exclusion and conservative paddocks was blocked from late 2010 to 16

spring/summer 2011. The conservative management started in late 2011. We resampled all sites 17

during the same period in 2011, 2012 and 2013. In each paddock we sampled the vegetation using 18

nine 1m2 permanent plots (systematically allocated in a 3x3 grid with 17 m between plots). In each 19

plot we surveyed all plant species that were present and estimated their cover using the decimal 20

scale of Londo (1976). We also estimated vegetation height in five points per sampling unit, and 21

estimated cover of bare soil, litter (aboveground dead biomass), rock outcrops and overall 22

vegetation cover per sampling unit. 23

We performed two experiments to estimate litter decomposition under the different 24

treatments using litter-bags (Wider & Lang 1982). In this procedure, we added a known mass of 25

standardized dry material into 10 x 10 cm bags made of green nylon mesh (1 mm2), with five 26

Page 129: abordagem quali-quantitativa e funcional de vegetação campestre ...

130

additional round perforations of 4 mm radius on each surface, which were fixed at ground level in 1

each sampling paddock using four specks. We chose this configuration for the litter-bags to 2

maximize camouflage in the environment, to minimize accidental loss of material and to allow the 3

entry of invertebrates from the soil macro fauna, which also contribute to decomposition (Swift et 4

al. 1979). After a determined period of time, we retrieved these bags, washed, dried and weighted 5

their remaining contents and estimated the decomposition rate after subtracting the final mass from 6

the initial mass. We used standardized materials in all sampling sites and paddocks, since we 7

searched for differences between decomposition rates related to the micro-environments created by 8

each treatment. Therefore, we did not take into account possible qualitative differences in litter (e.g. 9

decomposability) produced in each site/treatment. 10

In the first experiment we installed 32 litter-bags per paddock in five sites (logistic 11

problems made it impossible to install the experiment in one of the sites at the time), using dry 12

leaves of two broadly distributed plant species: Andropogon lateralis (Poaceae, 1g) and Eryngium 13

horridum (Apiaceae, 1.5g), with 16 bags each. The installation of this experiment took place in June 14

2011, and eight litter-bags of each species per paddock were removed six months later (December 15

2011), and the remaining litter-bags 18 months later (December 2012). The second experiment was 16

installed in all six sites in December 2012, and lasted six months (June 2013). As E. horridum and 17

A. lateralis materials presented extremely similar decomposition rates (based on the results from 18

the first experiment), in the second experiment we opted to substitute the later for cellulose filter 19

paper standard material which was expected to interact differently with the micro-environmental 20

conditions (e.g., not being affected by photodegradation; Vaieretti et al. 2010). We installed 16 21

litter-bags per paddock, eight containing dry leaves of E. horridum (1g) and eight with cellulose 22

filter paper (1g). 23

24

Data analysis 25

26

Page 130: abordagem quali-quantitativa e funcional de vegetação campestre ...

131

In our analyses we searched for differences between vegetation patterns among treatments 1

during four consecutive years of sampling. We estimated these differences considering species 2

composition and their relative abundances, species richness and diversity and functional aspects of 3

the community based on plant life forms. 4

We organized plant community data in a matrix containing average cover values of species 5

describing paddocks in each of the four years of sampling (matrix W). From matrix W we derived 6

matrices containing information of treatments and/or years separately, according to each analysis. 7

We classified plant species that were present in at least three paddocks in life form categories 8

(Table 1) proposed by Ferreira et al. (unpublished [Capítulo 2]). In this classification, life forms are 9

based on features such as habit, architecture, level of lignification and strategy of horizontal 10

occupation, which are responsive to shifts in management and good descriptors of vegetation 11

structure. Trait data were summarized in the binary matrix B of species described by life form 12

categories. To evaluate changes in dominance of life forms across the years of sampling and under 13

different treatments, we also generated a matrix T with community weighted mean traits by matrix 14

multiplication T = WB (Pillar et al. 2009). We used chord distance as dissimilarity measure 15

between sampling units in matrix W and Gower’s index (Podani 1999) in matrix B. 16

We calculated species diversity using Simpson’s index (Magurran & McGill 2011) and 17

functional diversity using Rao’s quadratic entropy (Botta‐Dukát 2005). We also calculated 18

functional redundancy using the method described in Pillar et al. (2013). Diversity indexes and 19

functional redundancy were calculated at the paddock level, and for every year of sampling. We 20

tested for correlations between diversity indexes and functional redundancy using correlation 21

analysis with permutation. To test for differences in indexes between years we used Repeated 22

Measures Analysis of Variance with permutation. We used ANOVA with permutation to evaluate 23

differences between treatments in the same sampling year, and between decomposition rates 24

between treatments. We restricted all permutations due to the blocked design of the experiment. 25

26

Page 131: abordagem quali-quantitativa e funcional de vegetação campestre ...

132

Results 1

2

Pairwise comparisons of species composition and abundance between years within each 3

treatment resulted in significant differences only in the exclusion, and only between the first and the 4

last two years (this difference vanishes when using only presence/absence data). Species richness 5

dropped sharply after the second year of exclusion, and differences between treatments became 6

significant after the third year, with higher values in the conventional treatment (Figure 1A). Plant 7

species diversity increased after the first year of exclusion, after which conventional and 8

conservative treatments maintained an increasing trend and exclusion decreased sharply. 9

Differences in diversity between the control and the two treatments were evident after the first year 10

of exclusion, became blurred in 2012 and were evident again in 2013 (Figure 1B). After four years, 11

species diversity was higher in the conservative treatment, contrasting with results for species 12

richness (Figure 1A). Considering the pooled data from the four years of monitoring, species 13

richness was significantly different only between the exclusion and conventional treatments 14

(P<0.05), and we found no significant differences considering species diversity. 15

Plant functional diversity showed little variation across four years in the control, whereas 16

we found opposite trends in the conservative and exclusion treatments: steady increase in the first 17

and sharp decrease in the latter (Figure 2A). Functional redundancy was overall lower in the 18

conventional treatment, increased linearly in the exclusion and decreased in the conservative after 19

the reintroduction of management in 2011 (Figure 2B). 20

The relative representativeness of life forms across years showed different trends in each 21

treatment. Cover of prostate plants remained roughly constant in the conventional treatment, 22

whereas it decreased in the conservative and exclusion (Figure 3A). Tussock cover was overall 23

lower in the conventional treatment (Figure 3B), whereas the conservative treatment promoted an 24

increase in cover of geophytes (Figure 3C). We found no clear pattern for annual plants (albeit a 25

small number of species with little cover only), although there is an indication of decreasing cover 26

Page 132: abordagem quali-quantitativa e funcional de vegetação campestre ...

133

in the exclusion after 2011 (Figure 3D). Forb cover increased after the first year of exclusion, but 1

the pattern disappeared in the following years (Figure 3E). Cover of shrubs and subshrubs increased 2

in the exclusion and remained unchanged in the conservative and conventional treatments (Figure 3

3F). 4

Relationships between functional diversity, species diversity and functional redundancy 5

varied between treatments (pooled data of four years of sampling). Functional diversity and species 6

diversity were significantly correlated in all treatments, although the magnitude of the correlation 7

decreased in the conservative and even more in the exclusion. The relationship between functional 8

redundancy and functional diversity followed the same pattern. Functional redundancy and species 9

diversity were strongly correlated in the conventional treatment, but the correlation dropped 10

considerably in the conservative and vanished in the exclusion (Figure 4). 11

Results of the first decomposition experiment showed no differences between treatments or 12

materials after the first six and 18 months of grazing exclusion (Jun 2011 – Dec 2012; results not 13

shown). However, results of the second experiment (Dec 2012 – Jun 2013) indicated differences in 14

decomposition rates between treatments and materials. Average values per treatment showed higher 15

decomposition rates of E. horridum in the conventional treatment (P<0.05; Figure 5A). After 16

separating these results by location of sampling sites (southern Pampa sites vs. northern Atlantic 17

Forest Sites), another pattern arose: decomposition rate of E. horridum was higher in conventional 18

paddocks from Pampa sites (P<0.01), whereas it varied little between treatments in Atlantic Forest 19

sites (Figure 5B). Decomposition rate of cellulose was less variable comparing treatments and sites, 20

although it was on average higher in Pampa sites (Figure 5C). Vegetation height increased sharply 21

in the conservative and exclusion treatments after one year, decreasing in the first and remaining 22

constant in the latter afterwards (Figure 6A). The percentage of dead biomass increased in the 23

exclusion and conservative managements across the years, and was significantly lower in the 24

conventional management after 2012 (Figure 6B). 25

26

Page 133: abordagem quali-quantitativa e funcional de vegetação campestre ...

134

Discussion 1

2

We aimed to investigate, by means of a controlled, randomized experiment, differences in 3

plant composition, diversity and ecosystem processes between grassland plant communities 4

submitted to different grazing regimes during four years. The first year of grazing exclusion 5

promoted a sharp increase in species diversity in the conservative (during the first year still without 6

the conservative management) and exclusion paddocks (Figure 1B). The cessation of grazing 7

produced a short-term decrease in the dominance of prostrate life forms, which was diluted among 8

other groups such as forbs, therophytes and ligneous plants (Figure 3). The conservative 9

management maintained the diversity levels achieved after one year of exclusion, although diversity 10

in the conventional management also increased with time and differences between conservative 11

management and exclusion were not significant after the third year of sampling. As expected, 12

species richness and diversity dropped significantly in the exclusion treatment after 2011, i.e., one 13

year of exclusion (Figure 1 A, B). In productive grasslands, the cessation of grazing is notably 14

followed by increased dominance of a few species (usually the less palatable ones) and the 15

associated decline of species adapted to grazing (e.g., Pucheta et al. 1998a; Pucheta et al. 1998b; 16

Cingolani et al. 2003; Vaieretti et al. 2010). Accordingly, we recorded a steady decline in cover of 17

prostrate plants (largely leaded by rhizomatous grass species) with concomitant increasing cover of 18

other life forms such as tussocks (Figure 3). Under increasing amounts of dead biomass and overall 19

taller vegetation (Figure 6), many species are outcompeted due to shading and tend to first decrease 20

in cover and eventually disappear (Tilman & Wedin 1991; Collins et al. 1995). Declines in species 21

richness in the exclusion treatment corroborate this hypothesis (Figure 1A). This competitive 22

exclusion also affected other life forms such as annual species (Figure 3D). Geophytes were evenly 23

represented among treatments prior to the experiment. Notwithstanding, after four years they were 24

more representative in the conservative treatment in comparison with the conventional and 25

exclusion, probably because they simply had their aerial parts removed by grazing in the first and 26

Page 134: abordagem quali-quantitativa e funcional de vegetação campestre ...

135

were outcompeted in the latter (Figure 3C). Shrubs and subshrubs, on the other hand, have benefited 1

from the exclusion of grazing (Figure 3F), and will probably play a key role on future shifts in 2

vegetation structure and diversity patterns in exclusion paddocks during the following years. 3

Although the conservative management also enabled different life forms to become more 4

representative in the community, woody plants were not benefited by this treatment, indicating that 5

no such shrub encroachment will take place under this management intensity. 6

Plant functional diversity (FD) was overall higher in the conventional treatment, although it 7

rose steadily after 2011 in the conservative and dropped linearly in the exclusion (Figure 2A). 8

Despite the overall higher FD in conventional paddocks prior to the experiment, the slight short-9

term increase and following maintenance of FD in this treatment may be attributed to the natural 10

spatial heterogeneity (Adler et al. 2001; McIntyre et al. 2003; McIvor et al. 2005) and differential 11

trait selection (Grime 2006; Diaz et al. 2007) promoted by grazing. We used plant life form 12

categories as traits to calculate FD, and the index we used relates to limiting similarity and niche 13

complementarity (Diamond 1975; Tilman et al. 1997; Wilson 1999) by measuring trait 14

dissimilarities among taxa taking their relative abundances into account (Botta‐Dukát 2005). 15

Therefore, higher values of FD mean higher distribution of abundances among life forms, i.e., less 16

dominance. The absence of grazing after 2011 promoted the homogenization of the community by 17

greatly reducing the contribution of prostrate species (Figure 3). The disappearance of the natural 18

patchiness was also reflected on average vegetation height, which doubled in the exclusion, 19

reflecting the increasing dominance of tussock grasses (Figure 6A). After 2011, when the 20

conservative management started, FD rose steadily and approached the control values, indicating 21

that the lighter grazing pressures under the simulated rotational grazing promoted, besides higher 22

species diversity (Figure 1B), more evenly distributed dominance of life forms (Figure 2A). Since 23

grazing animals stayed in a very restricted area in each conservative paddock during grazing events 24

(0.5 ha each), urine deposition may have also influenced heterogeneity by creating patches of high 25

productivity (Steinauer & Collins 1995; Steinauer & Collins 2001). In the exclusion treatment, 26

Page 135: abordagem quali-quantitativa e funcional de vegetação campestre ...

136

species became increasingly more redundant considering their life form through time (Figure 2B). 1

Coupled with decreasing species richness (Figure 1A), this indicates that species loss due to 2

management cessation may be taxonomically independent and very likely related to competitive 3

ability of selected species (Tilman 1984; Collins et al. 1995; Collins et al. 1998). In fact, theories 4

underlying plant-herbivore dynamics are largely based on tradeoffs between palatability and 5

competitive ability (Pacala & Crawley 1992), and although short-range dispersal and rapid 6

exploitation strategies may explain dynamics in annual plant communities (e.g., Bolker & Pacala 7

1999), the mechanisms underlying exploitation strategies in long-lived plant communities remain 8

obscure (Rees et al. 2001). 9

Considering the pooled data from the four years of sampling, the relation between species 10

diversity (D) and functional diversity (FD) was stronger in conventional paddocks, and 11

progressively weakened in conservative and exclusion paddocks (Figure 4). This indicates that 12

variation of FD through time tends to be independent of D with the cessation of management, 13

whereas both variables share similar variation patterns under considerably heavier grazing pressure. 14

The same relation held true between functional redundancy (FR) and FD: both variables are 15

progressively less correlated in the conservative and exclusion treatments. Paddocks with low 16

values of FD are dominated by few life forms, which would logically mean increased redundancy. 17

However, this relationship tends to weaken under lower intensity of disturbance. Finally, correlation 18

between FR and D is significant only in the conventional treatment, indicating that both variables 19

are independent under lighter or no grazing. Pillar et al. (2013), working in grazed grasslands in the 20

same region, showed that (at the plot scale) community stability was positively influenced by FR 21

and negatively influenced by grazing intensity. Our results showed that FR increased under low 22

(and under the absence of) grazing pressure (Figure 2B). Although our results represent mean 23

values per paddock and do not account for between-plot variation, we can assume that species that 24

are usually in ‘less palatable’ patches under grazing (Adler et al. 2001; McIntyre et al. 2003) are 25

progressively excluding more palatable plants and dominating exclusion paddocks with increasing 26

Page 136: abordagem quali-quantitativa e funcional de vegetação campestre ...

137

FR (see also Figure 3). Therefore, exclusion of management may be leading these communities to 1

increased resistance to grazing, considering resistance as the amount of external pressure needed to 2

cause a given amount of disturbance in the system (Carpenter et al. 2001). However, the same 3

communities are probably becoming less resistant to another common disturbance in grasslands: 4

fire. Light grazing pressures lead to accumulation of biomass (Figure 6B), and may also increase 5

fire intensity, extent and overall impact on grasslands and associated ecosystems (Bond & Keeley 6

2005; Fidelis et al. 2010). Moreover, cessation of management may be leading to decreased 7

resilience (Harrison 1979; Carpenter et al. 2001). The progressive removal of grazing-tolerant 8

species, the increase in woody species (Figure 3) and the shifting towards a closed and taller grass 9

canopy (Figure 6) may in time prevent these communities of shifting back to the original stage (i.e., 10

the conventional treatment), even if grazing is reintroduced. However, results obtained in the 11

conservative treatment suggest high resilience after a short period (1 year) of exclusion, since most 12

results were similar to the ones obtained in the control in the following years. 13

The absence of differences between treatments in the first decomposition experiment (Jun 14

2011 – Dec 2012) was probably time-related. Community structure was still in the process of 15

changing due to grazing exclusion and conservative management, and differences between micro-16

habitats and related organisms were still incipient. Although the second experiment was shorter (6 17

months, Dec 2012 – Jun 2013), the higher rates of decomposition of E. horridum in conventional 18

paddocks indicate that differences in vegetation structures promoted by different grazing pressures 19

influenced decomposition. Lighter grazing pressures promote more accumulation of dead biomass 20

and higher vegetation height (Figure 6), which probably decreased the rates of the litter 21

decomposition promoted by solar radiation (photodegradation; Pancotto et al. 2005). Higher 22

decomposition rates in grazed paddocks could also have been related to faeces and nitrogen 23

deposition by herbivores (Ruess & McNaughton 1987; Seagle et al. 1992; Bakker et al. 2004), a 24

factor that was absent in the exclusion. However, if such assumption was true, we would expect 25

even higher decomposition rates in the conservative treatment, where density of grazers per hectare 26

Page 137: abordagem quali-quantitativa e funcional de vegetação campestre ...

138

was much higher during grazing events, and so was deposition of faeces and urine. Also, 1

decomposition of cellulose did not follow the same pattern, providing further evidence that 2

differences between decomposition were most likely related to grassland canopy openness and the 3

UV radiation incidence, which primarily generate the molecular fragmentation of lignin 4

contributing to litter mass loss (Rutledge et al. 2010). When we separated decomposition results by 5

site, we found out that southernmost sites inserted in the Pampa biome were responsible for most of 6

the differences in decomposition of E. horridum between treatments (Figure 5B). When compared 7

with the Atlantic Forest biome, grasslands inserted in the Pampa biome are characterized by lower 8

canopies and less accumulation of biomass due to higher grazing pressures (Nabinger et al. 2000; 9

Nabinger et al. 2009). Accordingly, vegetation structure showed differences between sites from 10

different biomes prior to the implementation of the experiment we described here (Ferreira et al. 11

unpublished [Capítulo 1]). Recent findings found higher litter decomposition rates in sites with 12

more canopy openness in South Brazilian grasslands, which was probably due to increased 13

photodegradation (Podgaiski et al. 2014). 14

We reported here differences in grassland vegetation structure and diversity patterns 15

between paddocks that have been submitted to contrasting levels of grazing during four years. 16

Interruption of the disturbance regime caused drastic changes in community parameters by shifting 17

relationships of dominance, removing grazing-tolerant species and benefitting groups of species that 18

were controlled by grazing. Structural differences arising from management exclusion also 19

influenced litter decomposition. Our findings suggest that subtropical grasslands from South 20

America may be in an intermediate position in a gradient of resistance/resilience to grazing in 21

comparison with shortgrass steppes (Milchunas & Lauenroth 1993; Milchunas et al. 1998) and 22

tallgrass prairies (Coffin et al. 1996; Baer et al. 2000). Although four years may be considered a 23

short period of time to evaluate disturbance dynamics in comparison with studies carried out 24

elsewhere (e.g., Knapp et al. 1998; Rees et al. 2001; Turner et al. 2003), this is the first attempt to 25

Page 138: abordagem quali-quantitativa e funcional de vegetação campestre ...

139

do so in subtropical grasslands from South America. Also, all sites will be continuously monitored 1

as part of an ongoing LTER (PELD Campos Sulinos; CNPq 558282/2009-1). 2

3

Acknowledgements 4

5

We thank all farm owners that kindly allowed this work to be carried out in their properties, and all 6

local environmental authorities that allowed research in protected areas and helped us during 7

fieldwork. The first, second and third authors thank CAPES for scholarships. I. I. Boldrini thanks 8

CNPq for a research productivity grant. 9

Page 139: abordagem quali-quantitativa e funcional de vegetação campestre ...

140

Tables

Table 1. Life form categories used as binary traits to describe grassland plant communities. See

details in Ferreira et al. (unpublished [Capítulo 2]).

Life form category Acronym

Therophytes Th

Bulbous geophytes Bg

Rhizomatous geophytes Rg

Prostate rosette evergreens Pr

Decumbent evergreens De

Rhizomatous evergreens Rh

Stolonoiferous evergreens St

Solitary evergreen tussocks Te

Connected evergreen tussocks Ct

Evergreen forbs Ef

Erect rosette evergreens Er

Evergreen subshrubs Ss

Evergreen shrubs Sh

Succulent evergreens Su

Page 140: abordagem quali-quantitativa e funcional de vegetação campestre ...

141

Figures

Figure 1. Species richness and diversity along four years under three different grazing treatments:

conventional (conv), conservative (cons) and exclusion (exc). Mean values of six paddocks per

treatment/year. Different letters correspond to significant differences (P<0.05) within each year.

Page 141: abordagem quali-quantitativa e funcional de vegetação campestre ...

142

Figure 2. Functional diversity and redundancy along four years under three different grazing

treatments: conventional (conv), conservative (cons) and exclusion (exc). Mean values of six

paddocks per treatment/year. Different letters correspond to significant differences (P<0.05) within

each year.

Page 142: abordagem quali-quantitativa e funcional de vegetação campestre ...

143

Figure 3. Mean cover values of different life forms along four years under three different grazing

treatments: conventional (conv), conservative (cons) and exclusion (exc). Mean values of six

paddocks per treatment/year. See Table 1 for life form acronyms.

Page 143: abordagem quali-quantitativa e funcional de vegetação campestre ...

144

Figure 4. Correlation between functional diversity (Rao’s Q), species diversity Simpson’s D) and

functional redundancy (FR) in grassland communities under three different grazing treatments.

Mean values of six paddocks per treatment along four years of sampling (24 paddocks per analysis).

Functional redundancy was calculated as FR = D – Q (Pillar et al. 2013). P-values obtained using

restricted permutations.

Page 144: abordagem quali-quantitativa e funcional de vegetação campestre ...

145

Figure 5. Percentage of litter mass remaining after six months in litter bags containing two different

standardized materials (Eryngium horridum and cellulose paper), in grassland communities under

three different grazing treatments: conventional (conv), conservative (cons) and exclusion (exc). A.

Mean values of six sites (six paddocks per treatment). B,C. Mean values separated by biome (three

paddocks each) in which sites were inserted (AF = Atlantic Forest; P = Pampa).

Page 145: abordagem quali-quantitativa e funcional de vegetação campestre ...

146

Figure 6. Vegetation height and percentage of dead biomass along four years under three different

grazing treatments: conventional (conv), conservative (cons) and exclusion (exc). Mean values of

six paddocks per treatment/year. Different letters correspond to significant differences (P<0.05)

within each year.

Page 146: abordagem quali-quantitativa e funcional de vegetação campestre ...

147

References

Adler, P., Raff, D. & Lauenroth, W. 2001. The effect of grazing on the spatial heterogeneity of

vegetation. Oecologia 128: 465-479.

Araujo, P.I., Yahdjian, L. & Austin, A.T. 2012. Do soil organisms affect aboveground litter

decomposition in the semiarid Patagonian steppe, Argentina? Oecologia 168: 221-230.

Arnold, J.F. 1955. Plant life-form classification and its use in evaluating range conditions and

trend. Journal of Range Management 8: 176-181.

Augustine, D.J. & McNaughton, S.J. 1998. Ungulate effects on the functional species

composition of plant communities: herbivore selectivity and plant tolerance. The Journal of wildlife

management: 1165-1183.

Baer, S., Rice, C. & Blair, J. 2000. Assessment of soil quality in fields with short and long term

enrollment in the CRP. Journal of Soil and Water Conservation 55: 142-146.

Bakker, E., Olff, H., Boekhoff, M., Gleichman, J. & Berendse, F. 2004. Impact of herbivores on

nitrogen cycling: contrasting effects of small and large species. Oecologia 138: 91-101.

Bardgett, R.D. & Wardle, D.A. 2003. Herbivore-mediated linkages between aboveground and

belowground communities. Ecology 84: 2258-2268.

Behling, H. & Pillar, V. 2008. Vegetation and fire dynamics in southern Brazil during the late

Quaternary and their implication for conservation and management of modern grassland

ecosystems. In: Schröder, H. (ed.) Grasslands: Ecology, Management and Restoration, pp. 99-108.

Nova Science Publishers, New York.

Boldrini, I.I. 1997. Campos do Rio Grande do Sul: caracterização fisionômica e problemática

ocupacional. Boletim do Instituto de Biociências 56: 1-39.

Boldrini, I.I. & Eggers, L. 1996. Vegetação campestre do sul do Brasil: dinâmica de espécies à

exclusão do gado. Acta Botanica Brasilica 10: 37-50.

Page 147: abordagem quali-quantitativa e funcional de vegetação campestre ...

148

Boldrini, I.I., Eggers, L., Mentz, L.A., Miotto, S.T.F., Matzenbacher, N.I., Longhi-Wagner,

H.M., Trevisan, R., Schneider, A.A. & Setubal, R.B. 2009. Flora. In: Boldrini, I.I. (ed.)

Biodiversidade dos campos do Planalto das Araucárias, pp. 39-94. MMA, Brasília.

Bolker, B.M. & Pacala, S.W. 1999. Spatial moment equations for plant competition:

understanding spatial strategies and the advantages of short dispersal. The American Naturalist 153:

575-602.

Bond, W., Woodward, F. & Midgley, G. 2005. The global distribution of ecosystems in a world

without fire. New Phytologist 165: 525-538.

Bond, W.J. & Keeley, J.E. 2005. Fire as a global ‘herbivore’: the ecology and evolution of

flammable ecosystems. Trends in Ecology & Evolution 20: 387-394.

Botta‐Dukát, Z. 2005. Rao's quadratic entropy as a measure of functional diversity based on

multiple traits. Journal of Vegetation Science 16: 533-540.

Carpenter, S., Walker, B., Anderies, J.M. & Abel, N. 2001. From metaphor to measurement:

resilience of what to what? Ecosystems 4: 765-781.

Carrera, A., Bertiller, M. & Larreguy, C. 2008. Leaf litterfall, fine-root production, and

decomposition in shrublands with different canopy structure induced by grazing in the Patagonian

Monte, Argentina. Plant and soil 311: 39-50.

Chapin, F.S. 1993. Functional role of growth forms in ecosystem and global processes. In: Roy,

J., Ehleringer, J.R. & Field, C.B. (eds.) Scaling physiological processes: leaf to globe, pp. 287-312.

Academic Press, San Diego.

Cingolani, A.M., Cabido, M.R., Renison, D. & Solís Neffa, V. 2003. Combined effects of

environment and grazing on vegetation structure in Argentine granite grasslands. Journal of

Vegetation Science 14: 223-232.

Cingolani, A.M., Noy-Meir, I. & Díaz, S. 2005. Grazing effects on rangeland diversity: a

synthesis of contemporary models. Ecological Applications 15: 757-773.

Page 148: abordagem quali-quantitativa e funcional de vegetação campestre ...

149

Coffin, D.P., Lauenroth, W.K. & Burke, I.C. 1996. Recovery of vegetation in a semiarid

grassland 53 years after disturbance. Ecological Applications 6: 538-555.

Collins, S.L., Glenn, S.M. & Gibson, D.J. 1995. Experimental analysis of intermediate

disturbance and initial floristic composition: decoupling cause and effect. Ecology 76: 486-492.

Collins, S.L., Knapp, A.K., Briggs, J.M., Blair, J.M. & Steinauer, E.M. 1998. Modulation of

diversity by grazing and mowing in native tallgrass prairie. Science 280: 745-747.

Coughenour, M.B. 1991. Spatial components of plant-herbivore interactions in pastoral,

ranching, and native ungulate ecosystems. Journal of Range Management 44: 530-542.

Diamond, J.M. 1975. Assembly of species communities. In: Cody, M.L. & Diamond, J.M. (eds.)

Ecology and evolution of communities, pp. 342–444. Harvard University Press, Cambridge.

Diaz, S., Lavorel, S., McIntyre, S., Falczuk, V., Casanoves, F., Milchunas, D.G., Skarpe, C.,

Rusch, G., Sternberg, M. & NOY‐MEIR, I. 2007. Plant trait responses to grazing–a global

synthesis. Global Change Biology 13: 313-341.

Fidelis, A.T., Delgado Cartay, M.D., Blanco, C.C., Muller, S.C., Pillar, V.d.P. & Pfadenhauer,

J.S. 2010. Fire intensity and severity in Brazilian Campos grasslands. Interciencia: revista de

ciencia y tecnologia de america. Caracas. Vol. 35, n. 10 (Oct. 2010), p. 739-745.

Frank, D.A. & Evans, R.D. 1997. Effects of native grazers on grassland N cycling in

Yellowstone National Park. Ecology 78: 2238-2248.

Frank, D.A., Groffman, P.M., Evans, R.D. & Tracy, B.F. 2000. Ungulate stimulation of nitrogen

cycling and retention in Yellowstone Park grasslands. Oecologia 123: 116-121.

Frank, D.A. & McNaughton, S.J. 1993. Evidence for the promotion of aboveground grassland

production by native large herbivores in Yellowstone National Park. Oecologia 96: 157-161.

Grime, J. 1973. Control of species density in herbaceous vegetation. Journal of Environmental

Management 1: 151-167.

Grime, J. 1979. Plant Strategies and Vegetation Processes. In. New York: John Wiley.

Page 149: abordagem quali-quantitativa e funcional de vegetação campestre ...

150

Grime, J., Hodgson, J., Hunt, R., Thompson, K., Hendry, G., Campbell, B., Jalili, A., Hillier, S.,

Diaz, S. & Burke, M. 1997. Functional types: testing the concept in Northern England. In: Smith,

T., Shugart, H. & Woodward, F. (eds.) Plant functional types: their relevance to ecosystem

properties and global change, pp. 122-134. Cambridge University Press, Cambridge.

Grime, J.P. 2006. Trait convergence and trait divergence in herbaceous plant communities:

mechanisms and consequences. Journal of Vegetation Science 17: 255-260.

Güsewell, S., Jewell, P.L. & Edwards, P.J. 2005. Effects of heterogeneous habitat use by cattle

on nutrient availability and litter decomposition in soils of an Alpine pasture. Plant and soil 268:

135-149.

Hadar, L., Noy‐Meir, I. & Perevolotsky, A. 1999. The effect of shrub clearing and grazing on the

composition of a Mediterranean plant community: functional groups versus species. Journal of

Vegetation Science 10: 673-682.

Harrison, G.W. 1979. Stability under environmental stress: resistance, resilience, persistence,

and variability. American Naturalist: 659-669.

Hobbs, N.T. & Swift, D.M. 1988. Grazing in herds: when are nutritional benefits realized?

American Naturalist: 760-764.

Knapp, A.K., Briggs, J.M., Hartnett, D.C. & Collins, S.L. 1998. Grassland dynamics: long-term

ecological research in tallgrass prairie. Oxford University Press New York.

Londo, G. 1976. The decimal scale for relevés of permanent quadrats. Vegetatio 33: 61-64.

MacFadden, B.J. 2005. Diet and habitat of toxodont megaherbivores (Mammalia, Notoungulata)

from the late Quaternary of South and Central America. Quaternary Research 64: 113-124.

MacFadden, B.J. 1997. Origin and evolution of the grazing guild in New World terrestrial

mammals. Trends in Ecology & Evolution 12: 182-187.

Magurran, A.E. & McGill, B.J. 2011. Biological diversity: frontiers in measurement and

assessment. Oxford University Press Oxford.

Page 150: abordagem quali-quantitativa e funcional de vegetação campestre ...

151

McIntyre, B.S. & Tongway, D. 2005. Grassland structure in native pastures: links to soil surface

condition. Ecological Management & Restoration 6: 43-50.

McIntyre, S., Heard, K. & Martin, T.G. 2003. The relative importance of cattle grazing in

subtropical grasslands: does it reduce or enhance plant biodiversity? Journal of Applied Ecology 40:

445-457.

McIntyre, S., Lavorel, S. & Tremont, R. 1995. Plant life-history attributes: their relationship to

disturbance response in herbaceous vegetation. Journal of Ecology: 31-44.

McIvor, J.G., McIntyre, S., Saeli, I. & Hodgkinson, J. 2005. Patch dynamics in grazed

subtropical native pastures in south‐east Queensland. Austral Ecology 30: 445-464.

Milchunas, D., Lauenroth, W. & Burke, I. 1998. Livestock grazing: animal and plant

biodiversity of shortgrass steppe and the relationship to ecosystem function. Oikos: 65-74.

Milchunas, D., Sala, O. & Lauenroth, W.K. 1988. A generalized model of the effects of grazing

by large herbivores on grassland community structure. American Naturalist: 87-106.

Milchunas, D.G. & Lauenroth, W.K. 1993. Quantitative effects of grazing on vegetation and

soils over a global range of environments. Ecological Monographs 63: 327-366.

Nabinger, C., Ferreira, E., Freitas, A., Carvalho, P. & Sant’Anna, D. 2009. Produção animal com

base no campo nativo: aplicações de resultados de pesquisa. Campos sulinos: conservação e uso

sustentável da biodiversidade. Brasília: Ministério do Meio Ambiente: 175-198.

Nabinger, C., Moraes, A. & Maraschin, G.E. 2000. Campos in southern Brazil. In: Lemaire, G.,

Hodgson, J., Moraes, A., Carvalho, P.C.F. & Nabinger, C. (eds.) Grassland Ecophysiology and

Grazing Ecology, pp. 355-376. Cambridge University Press, United Kingdom.

Overbeck, G.E., Müller, S.C., Pillar, V.D. & Pfadenhauer, J. 2005. Fine‐scale post‐fire dynamics

in southern Brazilian subtropical grassland. Journal of Vegetation Science 16: 655-664.

Pacala, S. & Crawley, M. 1992. Herbivores and plant diversity. American Naturalist: 243-260.

Page 151: abordagem quali-quantitativa e funcional de vegetação campestre ...

152

Pancotto, V.A., Sala, O.E., Robson, T.M., Caldwell, M.M. & Scopel, A.L. 2005. Direct and

indirect effects of solar ultraviolet‐B radiation on long‐term decomposition. Global Change Biology

11: 1982-1989.

Pickett, S.T. 1985. The ecology of natural disturbance and patch dynamics. Academic press.

Pillar, V.D. 2009. Campos sulinos: Conservação e uso sustentável da biodiversidade. Ministério

do Meio Ambiente.

Pillar, V.D., Blanco, C.C., Müller, S.C., Sosinski, E.E., Joner, F. & Duarte, L.D. 2013.

Functional redundancy and stability in plant communities. Journal of Vegetation Science 24: 963-

974.

Pillar, V.D., Duarte, L.d.S., Sosinski, E.E. & Joner, F. 2009. Discriminating trait‐convergence

and trait‐divergence assembly patterns in ecological community gradients. Journal of Vegetation

Science 20: 334-348.

Podani, J. 1999. Extending Gower's general coefficient of similarity to ordinal characters.

Taxon: 331-340.

Podgaiski, L.R., Goldas, C.d.S., Ferrando, C.P.R., Silveira, F.S., Joner, F., Overbeck, G.E.,

Mendonça-Jr., M.d.S. & Pillar., V.D.P. 2014. Burning effects on detritivory and litter decay in

Campos grasslands. Austral Ecology. Austral Ecology.

Porto, A. 1954. História das missões orientais do Uruguai. Selbach, Porto Alegre.

Pucheta, E., Cabido, M., Díaz, S. & Funes, G. 1998a. Floristic composition, biomass, and

aboveground net plant production in grazed and protected sites in a mountain grassland of central

Argentina. Acta Oecologica 19: 97-105.

Pucheta, E., Vendramini, F., Cabido, M. & Díaz, S. 1998b. Estructura y funcionamiento de un

pastizal de montaña bajo pastoreo y su respuesta luego de su exclusión. Revista de la Facultad de

Agronomía 103.

Page 152: abordagem quali-quantitativa e funcional de vegetação campestre ...

153

Quadros, F.L.F., Cruz, P., Theau, J.P., Jouany, C., Duru, M., Carvalho, P.C.F. & Frizo, A. 2006.

Uso de tipos funcionais de gramíneas como alternativas de diagnóstico da dinâmica e do manejo de

campos naturais. In: 42 Reunião Anual da Sociedade Brasileira de Zootecnia, João Pessoa.

Raunkiaer, C. 1934. The life forms of plants and statistical plant geography. Claredon, Oxford.

Rees, M., Condit, R., Crawley, M., Pacala, S. & Tilman, D. 2001. Long-term studies of

vegetation dynamics. Science 293: 650-655.

Ruess, R. & McNaughton, S. 1987. Grazing and the dynamics of nutrient and energy regulated

microbial processes in the Serengeti grasslands. Oikos: 101-110.

Rutledge, S., Campbell, D.I., Baldocchi, D. & Schipper, L.A. 2010. Photodegradation leads to

increased carbon dioxide losses from terrestrial organic matter. Global Change Biology 16: 3065-

3074.

Sala, O.E. 1988. The effect of herbivory on vegetation structure. In: Werger, M.J.A., van der

Aart, P.J.M., During, H.J. & Verboeven, J.T.A. (eds.) Plant form and vegetation structure, pp. 317-

330, The Hague.

Seagle, S.W., McNaughton, S. & Ruess, R.W. 1992. Simulated effects of grazing on soil

nitrogen and mineralization in contrasting Serengeti grasslands. Ecology: 1105-1123.

Semmartin, M., Garibaldi, L.A. & Chaneton, E.J. 2008. Grazing history effects on above-and

below-ground litter decomposition and nutrient cycling in two co-occurring grasses. Plant and soil

303: 177-189.

Senft, R., Coughenour, M., Bailey, D., Rittenhouse, L., Sala, O. & Swift, D. 1987. Large

herbivore foraging and ecological hierarchies. BioScience 37: 789-795.

Soussana, J.-F. 2009. Os desafios da ciência das pastagens européias são relevantes para os

Campos Sulinos. In: Pillar, V.D., Müller, S.C., Castilhos, Z.M.S. & Jacques, A.V.A. (eds.) Campos

Sulinos: Conservação e Uso Sustentável da Biodiversidade, pp. 331-344. Ministério do Meio

Ambiente, Brasília.

Page 153: abordagem quali-quantitativa e funcional de vegetação campestre ...

154

Steinauer, E. & Collins, S. 1995. Effects of urine deposition on small-scale patch structure in

prairie vegetation. Ecology: 1195-1205.

Steinauer, E.M. & Collins, S.L. 2001. Feedback loops in ecological hierarchies following urine

deposition in tallgrass prairie. Ecology 82: 1319-1329.

Swift, M.J., Heal, O.W. & Anderson, J.M. 1979. Decomposition in terrestrial ecosystems.

University of California Press, Berkeley, California, USA.

Throop, H.L. & Archer, S.R. 2007. Interrelationships among shrub encroachment, land

management, and litter decomposition in a semidesert grassland. Ecological Applications 17: 1809-

1823.

Tilman, D., Lehman, C.L. & Thomson, K.T. 1997. Plant diversity and ecosystem productivity:

theoretical considerations. Proceedings of the National Academy of Sciences 94: 1857-1861.

Tilman, D. & Wedin, D. 1991. Dynamics of nitrogen competition between successional grasses.

Ecology 72: 1038-1049.

Tilman, G.D. 1984. Plant dominance along an experimental nutrient gradient. Ecology: 1445-

1453.

Turner, M.G., Collins, S.L., Lugo, A.L., Magnuson, J.J., Rupp, T.S. & Swanson, F.J. 2003.

Disturbance dynamics and ecological response: the contribution of long-term ecological research.

BioScience 53: 46-56.

Vaieretti, M.V., Cingolani, A.M., Harguindeguy, N.P. & Cabido, M. 2013. Effects of differential

grazing on decomposition rate and nitrogen availability in a productive mountain grassland. Plant

and soil 371: 675-691.

Vaieretti, M.V., Cingolani, A.M., HARGUINDEGUY, N.P., Gurvich, D.E. & Cabido, M. 2010.

Does decomposition of standard materials differ among grassland patches maintained by livestock?

Austral Ecology 35: 935-943.

Page 154: abordagem quali-quantitativa e funcional de vegetação campestre ...

155

Verhoef, H.A., Verspagen, J.M. & Zoomer, H.R. 2000. Direct and indirect effects of ultraviolet-

B radiation on soil biota, decomposition and nutrient fluxes in dune grassland soil systems. Biology

and Fertility of Soils 31: 366-371.

Wider, R.K. & Lang, G.E. 1982. A critique of the analytical methods used in examining

decomposition data obtained from litter bags. Ecology: 1636-1642.

Wilson, J.B. 1999. Assembly rules in plant communities. In: Weiher, E. & Keddy, P. (eds.)

Ecological assembly rules: Perspectives, Advances, Retreats, pp. 130–164. Cambridge University

Press, Cambridge.

Page 155: abordagem quali-quantitativa e funcional de vegetação campestre ...

156

Considerações finais

Trabalhar com vegetação campestre é muito bom. Decidi iniciar esta breve seção final da

tese com esta frase porque ela resume minha experiência ao longo do doutorado (e do mestrado

também). Eu gosto do que faço, seja em campo de bombacha, bota e chapéu, seja de bermuda e

chinelo no computador, brincando no R ou no Multiv. É claro que a companhia ajuda muito: tenho

a sorte de trabalhar ao lado de pessoas (na maioria) ótimas. Trabalhar em um projeto grande como o

PELD Campos Sulinos foi uma experiência extremamente desafiadora e recompensadora. Tenho

orgulho de ter participado do início deste projeto, que tem a ambiciosa meta de estabelecer sítios de

pesquisa permanentes (fato ainda raro no Brasil, sobretudo em ecossistemas campestres).

Infelizmente não há espaço físico em uma tese de doutorado para inlcuir tudo o que foi feito ao

longo de quatro anos (nem tempo para escrever, na verdade). Sei que muitos doutorandos são

extremamente focados nas suas teses. Admiro estes colegas, mas eu não sou nada focado. Participei

de diversos projetos paralelos nestes quatro anos, especialmente junto ao PPG Ecologia da UFRGS,

no Laboratório de Ecologia Quantitativa (ECOQUA, para os íntimos). Provavelmente esta tese teria

sido mais organizada, consistente e com mais capítulos se eu não tivesse feito isso, mas não me

arrependo de absolutamente nenhum projeto paralelo ou conversa informal de quatro horas com

algum colega sobre uma análise legal. Na verdade fui orientado a ser assim (obrigado, Ilsi).

Também é necessário ressaltar aqui o aspecto multidisciplinar de se trabalhar com

vegetação campestre. Diferentemente de ecossistemas florestais, os campos bem conservados não

estão em unidades de conservação, ou em áreas extremamente inacessíveis. Campos bem

conservados estão nas propriedades rurais que bravamente mantêm a pecuária sobre campo nativo

em seus mosaicos de produção. Trabalhar nestas propriedades é sair do computador na sala da

Universidade e voltar no tempo para um mundo de estradas de chão, gado, cavalos e gaúchos (em

extinção) que falam uma língua entre o português e o espanhol. É um processo de imersão cultural

que pesquisadores das áreas sociais invejariam. O biólogo ‘típico’ não tem o costume de trabalhar

Page 156: abordagem quali-quantitativa e funcional de vegetação campestre ...

157

neste meio, que geralmente é mais ligado ao agrônomo, veterinário ou zootecnista. O contato com

estes profissionais leva à reflexão sobre conceitos tidos como óbvios nas ciências biológicas. Esta

interação deveria ser mais estimulada em programas de pós-graduação, pois leva a um inegável

crescimento mútuo advindo de diferentes experiências e visões da diversidade biológica. Trabalhos

de várias partes do mundo (incluindo aqui as nossas singelas contribuições) apontam a importância

do manejo da vegetação campestre na manutenção da sua diversidade em diversos níveis. Apesar

disso, ainda há profissionais das áreas biológicas que consideram áreas de campo manejadas com

fogo e/ou gado como ambientes degradados, e que a conservação passa obrigatoriamente pela

exclusão total de intervenção humana. Esta visão dogmática desaparece em dez minutos de

conversa com um ecólogo de pastagens, mas também com um pecuarista bem informado ou um

peão de estância com 40 anos de lida, pois ambos veem na prática o que acontece com uma área

campestre não manejada.

Acredito que o conjunto dos dados apresentados nesta tese forme uma contribuição

consistente, porém pontual, ao estudo dos ecossistemas campestres do Rio Grande do Sul. O legado

mais relevante destes quatro anos de trabalho talvez seja o conjunto bruto de dados obtidos em

quatro anos consecutivos de levantamento da vegetação sob diferentes manejos em um experimento

controlado. Estes dados serão a referência de comparação para os dados obtidos nos anos

subsequentes deste projeto de longa duração.